How to bend wood like this - Outdoor furniture project

[Alan Lightstone]

Looking for a new project to do. I've been spending a few months upgrading the workshop / fixing minor annoyances, and really need to actually build something.

I ran across this in a magazine ad.

516165

The furniture is supposedly solid teak, though I don't think that will be in the cards for mine.

These are serious bends in the backs of the chairs and couch. I didn't think teak was bendable like that.

So my question is how can I build something like this?
I'm not terrible fond of steam bending, as in my previous projects springback has been inconsistent.
Bend laminations is possible, though hiding it on a veneer top with the top curved also seems difficult.

So what's the best way to do this?
And, of course, wood choice - it has to live in brutal Florida weather. The area is covered, but it does get afternoon sun. I've built outdoor furniture for here before, but the though of refinishing in a few years always makes me cringe. Since I don't see small slats here on the design, I think in a few years that would be easier than, for example, my Titanic deck chair which is a refinishing nightmare.

[Lee Schierer]

It is probably done by gluing up thin layers of wood, similar to plywood, but with the grain all running parallel.

[John TenEyck]

What Lee said. If you look closely at the left chair you can see that the out layer is not a single piece. It's veneer that has been seamed. That supports Lee's hypothesis that it was as a glued up laminate. One could still argue that it's solid wood.

You don't need to hide the layers where they show at the top, but if you want to, you could make pieces maybe 3/4" think and cut them to fit between the outer plies of veneer when you bend them into the form. Use three pieces for a more continuous grain, just like is done when building a chair with solid wood.

In any case, use quality epoxy and they will last a long, long time.

As for refinishing, I've said this before, the only way to avoid it is to never finish or to cover the furniture when not in use. I cover mine and haven't done any maintenance to the Epifanes varnish in more than 5 years. Before that, it was every year or two.

John

[Derek Cohen]

To bend solid oak like that you require a serious bending press ...

https://www.youtube.com/watch?v=R_cyx7ti1B4&ab_channel=EngelsCoachShop

Regards from Perth

Derek

[Alan Lightstone]

What Lee said. If you look closely at the left chair you can see that the out layer is not a single piece. It's veneer that has been seamed. That supports Lee's hypothesis that it was as a glued up laminate. One could still argue that it's solid wood.

You don't need to hide the layers where they show at the top, but if you want to, you could make pieces maybe 3/4" think and cut them to fit between the outer plies of veneer when you bend them into the form. Use three pieces for a more continuous grain, just like is done when building a chair with solid wood.

In any case, use quality epoxy and they will last a long, long time.

As for refinishing, I've said this before, the only way to avoid it is to never finish or to cover the furniture when not in use. I cover mine and haven't done any maintenance to the Epifanes varnish in more than 5 years. Before that, it was every year or two.

John
Ah, hadn't seen the seam on the left chair/couch. That also might have been done to decrease the length of the board needed to make that couch.

I don't see a seam on the chair in the middle.

I thought this had to be bent lamination, but they seem to have done a very good job at the veneer seams on the top of the arms.

Not sure if I understand the 3/4" thick wood use.

[Phil Gaudio]

One of my earlier posts on curved work: should give you some idea of how it is done.
https://sawmillcreek.org/showthread.php?306584-Bowfront-Console-Table-Bent-Lamination&p=3290471&highlight=#post3290471

[Jim Becker]

Watch Darren Oatts on the 'Tube for the bending techniques.

[Jamie Buxton]

There's another way to build the curved shape, and it does not involve bending the wood. You bricklay the rough shape, and then bandsaw it to the final shape. With no bending involved, the construction is easier, and the glue lines have no stress on them.

[Jim Becker]

There's another way to build the curved shape, and it does not involve bending the wood. You bricklay the rough shape, and then bandsaw it to the final shape. With no bending involved, the construction is easier, and the glue lines have no stress on them.
True, although in the designs that OP shows in the photo that would not provide as clean a look to the furniture components because the glue lines will become a pattern as you build up layers. Those components are, my guess, 6-8" tall. Now an interesting option here would be to build the "core" of the components like you state and then bend a "thick veneer" around them, inside and out.

[Alan Lightstone]

There's another way to build the curved shape, and it does not involve bending the wood. You bricklay the rough shape, and then bandsaw it to the final shape. With no bending involved, the construction is easier, and the glue lines have no stress on them.

That's really interesting, Jamie. Never heard of that. Any Youtube videos on that, or threads here.

[Alan Lightstone]

Watch Darren Oatts on the 'Tube for the bending techniques.
Funny, a quick search for "Darren Oatts" on YouTube gives you lots of Hall and Oates videos.

[Jim Becker]

Funny, a quick search for "Darren Oatts" on YouTube gives you lots of Hall and Oates videos.
LOL I fat-fingered the spelling

https://www.youtube.com/@darrenoatesfinefurniture3648

[Alan Lightstone]

LOL I fat-fingered the spelling

https://www.youtube.com/@darrenoatesfinefurniture3648
Thanks, Jim. LOL.

Very impressive setup for steam bending wood. Best I've ever seen.

Now next question then. What are good outdoor woods that are also good for steam bending?
And how in the world to get springback to stay the same in each piece. Leave them bent in the form for a long time?

[Michael Burnside]

There's another way to build the curved shape, and it does not involve bending the wood. You bricklay the rough shape, and then bandsaw it to the final shape. With no bending involved, the construction is easier, and the glue lines have no stress on them.

This is how I have done it before. Bandsaw and CNC both work and it looks quite nice IMHO and is wicked strong. I'm actually working on a table soon that will utilize this technique again.

In Alan's original picture there is at least one seam on the sofa so I'm not sure if this is a similar trick with half-rounds or if it's a veneer of sorts.

[Jimmy Harris]

I see the seam in the wood on the side of the couch, but not the top. And with the seam occurring deep in the bend, and appearing to be a butt joint, my guess is that we're looking at veneers. So it's probably the bricklayed wood method, or something similar, and then veneered over the sides.

You could also try to make a bunch of kerf cut on the inside of the bend. I've seen people make some pretty complex shapes like that. But you wouldn't get the strength of the bricklayed method, but it might still be strong enough. You would, however, not have to use veneers. And you could call it solid wood, if that matters. You might still have to steam it to get it to bend that much, and might need to use green wood, depending on the species, which can cause a whole bunch of other issues.

[Mel Fulks]

Alan, that should have been “to my esteamed Friend”.

[Jim Becker]

Thanks, Jim. LOL.

Very impressive setup for steam bending wood. Best I've ever seen.

Now next question then. What are good outdoor woods that are also good for steam bending?
And how in the world to get springback to stay the same in each piece. Leave them bent in the form for a long time?
White oak (preferably air dried), Mahogany, etc. You do the bend for the layers twice; with steam first and left to dry in the form and then glued and clamped in the form to final. You need to do this as a bent lamination, not full thickness.

[John TenEyck]

True, although in the designs that OP shows in the photo that would not provide as clean a look to the furniture components because the glue lines will become a pattern as you build up layers. Those components are, my guess, 6-8" tall. Now an interesting option here would be to build the "core" of the components like you state and then bend a "thick veneer" around them, inside and out.

That technique is sometimes used for building drawers with curved fronts.

John

[John TenEyck]

Ah, hadn't seen the seam on the left chair/couch. That also might have been done to decrease the length of the board needed to make that couch.

I don't see a seam on the chair in the middle.

I thought this had to be bent lamination, but they seem to have done a very good job at the veneer seams on the top of the arms.

Not sure if I understand the 3/4" thick wood use.

This is what I meant:

https://lh3.googleusercontent.com/pw/AP1GczM3nRynNBC-f9x17PWr8_D6OwSOY1dlVyqIeAbP15t57P9466uFYq3cFU9srO ghoYWYzTXquZ0tDwE1RYBnJ4hrkHTFVut5DMkvyLsDWiJxvDcR FIvfQfO1VdaIzssJZr3r2QPVIt_nwOLjx7z1rDoGXQ=w1460-h604-s-no?authuser=1

The ring sits on top of the not shown inner plies. The ring is made from solid wood and preferably is made from 3 or more segments to keep the grain running roughly parallel with the curve.

John

[Alan Lightstone]

This is what I meant:

https://lh3.googleusercontent.com/pw/AP1GczM3nRynNBC-f9x17PWr8_D6OwSOY1dlVyqIeAbP15t57P9466uFYq3cFU9srO ghoYWYzTXquZ0tDwE1RYBnJ4hrkHTFVut5DMkvyLsDWiJxvDcR FIvfQfO1VdaIzssJZr3r2QPVIt_nwOLjx7z1rDoGXQ=w1460-h604-s-no?authuser=1

The ring sits on top of the not shown inner plies. The ring is made from solid wood and preferably is made from 3 or more segments to keep the grain running roughly parallel with the curve.

John
Really interesting. Never saw that before. Thanks, John.

[Alan Lightstone]

White oak (preferably air dried), Mahogany, etc. You do the bend for the layers twice; with steam first and left to dry in the form and then glued and clamped in the form to final. You need to do this as a bent lamination, not full thickness.
That's a great idea, Jim. Never seen that described before in any articles/posts on bending. Definitely going to try that approach.

[Alan Lightstone]

To bend solid oak like that you require a serious bending press ...

https://www.youtube.com/watch?v=R_cyx7ti1B4&ab_channel=EngelsCoachShop

Regards from Perth

Derek
Impressive. Petrifying, but impressive.

[Jim Becker]

That's a great idea, Jim. Never seen that described before in any articles/posts on bending. Definitely going to try that approach.
That's the method that is shown in many of the videos on the YouTube channel I referred you too. Almost all of Darren's projects involved bent lamination.

[Alan Lightstone]

LOL I fat-fingered the spelling

https://www.youtube.com/@darrenoatesfinefurniture3648
I can't find where he has listed the parts / equipment he uses for that pulley/form system. It really is the best I've ever seen. I'd love to duplicate that for furniture in the future.

[Alan Lightstone]

That's the method that is shown in many of the videos on the YouTube channel I referred you too. Almost all of Darren's projects involved bent lamination.
I've seen bent lamination. But never a combination of steam bending and bent lamination. Seems to correct a number of issues with just doing steam bending (especially springback).

[Maurice Mcmurry]

Some big manufacturers use ammonia gasification for things like that.
Lets talk about Ammonia [Archive] - Sawmill Creek Woodworking Community (https://sawmillcreek.org/archive/index.php/t-142975.html)

Steam bending Pryor to lamination works well and is used in making banjo rims and boat ribs. These chair back legs are done that way too.

516182 516183 516181

[John TenEyck]

Really interesting. Never saw that before. Thanks, John.

Not my idea; I've seen it done by Michael Fortune, I think.

John

[Andrew Hughes]

Have you thought about cypress for that outdoor project. Should be plenty of it in Florida the cypress we have on the west coast everyone calls cedar its pretty easy to work and bends great. Not sure about particular species out there is it Bald cypress. Ive used it once though it was interesting wood almost waxing if I remember right.
Its a nice looking design.
Good Luck

[Jim Becker]

I can't find where he has listed the parts / equipment he uses for that pulley/form system. It really is the best I've ever seen. I'd love to duplicate that for furniture in the future.
That's because he doesn't. But it you do some screen scrapes, it should be something you can figure out.

[Steve Demuth]

Steam bending and then gluing the laminations is a great technique where you need small radius bends relative to the mass of the final form. I use it to create rims for basket construction.

516204516205516206

The radius on the trug rims is only an inch and a half or so on the inside of the curves. Wood choice also matters, of course. These are made from 1/8" lams of red elm - elm is stringy, fibrous wood that takes heat bending extremely well. Not suitable for outdoor furniture, but brilliant for stuff like this.

[Alan Lightstone]

That's because he doesn't. But it you do some screen scrapes, it should be something you can figure out.
Pity. I've been trying to figure out a few of the parts, but not something I usually work with.

So I guess that's a cable winch. I'm online searching for winch part names.
516208

And, the black piece in the middle that looks like two pulley wheels that the wire goes through and then is attached to both sides of the wood to tighten it. And the red piece that the load hook grabs on to that also seems to be a pulley. I think that's a "Snatch Block".

That system looks great. I'd love to know the name of those parts so I could order them. It really looks to be the best setup I've seen:
516207

[Steve Demuth]

He's got a ratcheting winch to provide the force to the cable. The orange and red pulley devices he's using to redirect the force are called snatch blocks. Easy to find in a good hardware store, or even a bad one like Amazon.

[Alan Lightstone]

He's got a ratcheting winch to provide the force to the cable. The orange and red pulley devices he's using to redirect the force are called snatch blocks. Easy to find in a good hardware store, or even a bad one like Amazon.
Thanks, Steve. Never worked with this stuff, but have always wanted to build some furniture with curves and I love this system. Have to start buying and building now. And, of course, wood choice and obtaining good wood for four chairs and a couch. Quite a project for me. Likely to take a good year. But I think I'll love the finished project. Hopefully SWMBO too.

[Curt Harms]

That's really interesting, Jamie. Never heard of that. Any Youtube videos on that, or threads here.

I remember a New Yankee Workshop episode using that technique. I think it was building something like outdoor arches for fencing.

[jack duren]

Regardless of the pictures , you’ll have to decide how you’ll make the bends.

It doesn’t matter about the picture..

it doesn’t matter how they did it, just how you do…

[Alan Lightstone]

So I got a very good look at the original, and took a bunch of pictures. There are actually a few seams, so bending will be much easier. Now have a good idea where to start, and make a good game plan.
516220
And here's a picture of the seam in the back:
516221

Now still have lots of practice to do with steam lamination followed by bent lamination, and need to experiment to see how thick the pieces can be that can be bent successfully, but should be fun.

I ordered the winch, and assorted snatch blocks. Need to get wire, caribiners, etc... Should be an interesting adventure.

Lots and lots and lots of duplicate pieces needed.

[John TenEyck]

I see 5 joints in the back in the top photo. It wasn't steam bent at all. It was made from solid sections that were joined and then sawn into the desired curve.

John

[Jamie Buxton]

That's really interesting, Jamie. Never heard of that. Any Youtube videos on that, or threads here.

Here's a youtube about it. The guy makes it seem a little more challenging than what it actually is. https://www.youtube.com/watch?v=pJL7l_QtNv8

Notice that there's no bending of the wood. And there's no end-grain glue joints. In fact, all the glue faces are big facegrain-to-facegrain areas, so the resulting curved piece of wood is very strong.

[Alan Lightstone]

I see 5 joints in the back in the top photo. It wasn't steam bent at all. It was made from solid sections that were joined and then sawn into the desired curve.

John
That would explain why they advertise it as solid teak. Probably need at least 8/4 teak to do that. Definitely not in my budget.

[Jim Becker]

The winch shown is a simple, inexpensive boat trailer winch...available at Harbor Freight. (I use one on my utility trailer) The design of that fixture provides for even pressure from both sides with the wyw connection to the winch.

[Alan Lightstone]

The winch shown is a simple, inexpensive boat trailer winch...available at Harbor Freight. (I use one on my utility trailer) The design of that fixture provides for even pressure from both sides with the wyw connection to the winch.
Great. Parts ordered. Building to come. I'm going to try the combo of steam bending and bent lamination, even though the original seems to be built from solid wood. I like learning new techniques, plus not thrilled with all the joinery that would be involved with solid wood.
The project will be a challenge (plus need to find custom cushions for them eventually), but challenges are so much more satisfying than easy stuff.

[Jim Becker]

By making the fixture modular, you can have "different bends" available with the same mechanism. Some of his fixtures use only clamps, BTW. Pay attention to how he takes the time to align the laminations vertically, too, which makes for less cleanup after a component comes out of the fixture.

[John TenEyck]

That would explain why they advertise it as solid teak. Probably need at least 8/4 teak to do that. Definitely not in my budget.

I think 6/4 would be thick enough. Still pricey in teak, but there are lots of other suitable woods. Iroko comes to mind, as well as mahogany. Both should be available for $15/bf or less.

John

[Tom Bender]

That rig looks like a crossbow. If the lamination slips it could throw a clamp at you. Take precautions.

The more basic method of clamping around a form is safer.

Another method uses inner and outer forms pulled together.

Please keep us posted.

[Alan Lightstone]

That rig looks like a crossbow. If the lamination slips it could throw a clamp at you. Take precautions.

The more basic method of clamping around a form is safer.

Another method uses inner and outer forms pulled together.

Please keep us posted.
Yes, I thought the same thing. That rig does have my attention, but I am planning on using something similar.

Probably going to order 8/4 QS White Oak today. I doubt they'll have air-dried, but I'll ask.

The finished piece needs to be 1-1/2" thick. I wonder what the maximum thickness I'll be able to bend - decrease the number of plies of the lamination. I'm assuming that steam bending them first will allow greater thickness per piece than just using bent lamination.

Gluing up the curved pieces may be no fun, but we'll see.

Michael Fortunes method of steam bending - taking the piece out and bending in both directions to break the lignin bonds is interesting. https://www.finewoodworking.com/2007/03/15/free-form-steam-bending-technique
Don't think I'm going to try it, though he knows FAR more about this than I do.

It actually looks like ordering the cushions for these will cost far more than the chair itself. That curved back pattern isn't made by most companies I can find. The originals are about $1K per chair from the company that builds these chairs. Ouch!!!

[Tom Bender]

Try an automotive or boat upholsterer.

Last time I needed to bend some oak I went to a mill that mills anything he can get cheap and makes pallet wood and crane mats. He sold me very green oak for a song. Bends best the same day it's milled.

[William Hodge]

For small stuff, a form and 2" truck straps work well. The laminations need to be thin enough to easily bend on the required radius. Padding inside and outside the work piece needs to be adequate to spread the force of the clamps evenly.

[Alan Lightstone]

Wood arrived yesterday. Busy making the forms now. Using 3/4"MDF. OMG the weight of 8 of these on top of each other to accommodate the 5-1/2" wide wood that needs to be bent will require help, to say the lease.

I made the curved portion of the form out of 1/2" MDF using a Carter Accuright circle cutting jig that I had on my bandsaw. What a PIA. Shoulda just made a trammel with my router. Probably would have come out a little smoother too, though mine will work. Killing me the waste of MDF when I'm done. Not that I ever use MDF, but still taking a ton to make the bending form.

I couldn't lift or transport 4x8 MDF sheets, so I'm going to make this out of 2x4 MDF sheets and domino them together.

Now I need to build a larger steaming box than I have, and probably get a larger steam generator. Lots and lots of work to do. Any good suggestions for steam generators. Somewhere I read that wallpaper steamers, whatever they are, are good.

[Jim Becker]

When you make the layers for the forms, make one that's "da bomb" and use that to template route the remaining layers for a perfect form.

[jack duren]

So are you using smaller strips or trying to bend solid?

[Alan Lightstone]

So are you using smaller strips or trying to bend solid?
I think it will have to be smaller strips. I can't imagine being able to bend 1-1/2" white oak. And I couldn't get air dried, so it's kiln dried.

Still not sure the maximum thickness I can bend successfully. Obviously, the thicker the laminations, the less work I'll have to do.

Any guesses as to whether or not 1/2" white oak can be bent successfully? 3/4"??

[Alan Lightstone]

When you make the layers for the forms, make one that's "da bomb" and use that to template route the remaining layers for a perfect form.
I agree. That's the game plan. I'll bandsaw outside the line by 1/8" or so, then pattern rout the remainder.

[jack duren]

You’ll probably have go to 3/16. Depending on width you may need 1/8..

Air or kiln dried makes a difference

[Alan Lightstone]

You’ll probably have go to 3/16. Depending on width you may need 1/8..

Air or kiln dried makes a difference

Ouch. Lots and lots and lots of laminations.

[jack duren]

Is what it is..1/8x 6=3/4 bend all at once..

[Jim Becker]

You’ll probably have go to 3/16. Depending on width you may need 1/8..

Air or kiln dried makes a difference
I agree with this.

[Alan Lightstone]

I need 1-1/2" final thickness, so at minimum 8 pieces, at most 12. Lotta gluing of curved parts. Oh well...

[Jim Becker]

Yes, it requires a lot of layers, but the end result, if you are careful, will be outstanding.

[Alan Lightstone]

Making the forms out of 3/4" mdf. Had several people helping load the MDF and the truck I had to hire to move the 4x4 half sheets of MDF. Lots of work just to get 8 4x4 sheets to the workshop.

OMG the amount of MDF dust that is generated by pattern routing the edge even with my Festool router hooked up to its dust collector, both Jet Air cleaners going, and my 3M P100 respirator tightly around my face.

Staggering amount of dust EVERYWHERE. I don't know how people can work with this stuff. I very rarely do, and now I remember why. Working with it without serious respirators is such a recipe for lung disease. Even after vacuuming everything I can see after each rout, I will be finding deposits on everything and vacuuming more and more for weeks. Ouch.

[mike stenson]

Yea, when I have to use it I use a N95 rated respirator and turn on the extractor fan. When I'm done, I just leave the shop for a few hours and let the fines get blown outside (as much as possible)... and I'm still vacuuming in the end. I hate dealing with it.

[Ron Selzer]

yes it is very dusty to work with, did a lot with it at one time in my work life. get bunks of 3'x10' up to 5'x12'. was in shape when handling the 5'x12'. Always hated working with the imported mdf, liked domestic better.
Ron

[Alan Lightstone]

OK. Now the next problem. I have built 8 forms, to be attached to each other to make the required 5-1/2" height to bend the wood laminations against the jig. They have been rough cut on the bandsaw, and then using a pattern bit made as close as possible to the pattern:
516977

I have also on the drill press drilled a number of holes to use clamps to hold the wood against the bending jig:
516978

My present problem is how to drill all those holes in the 8 pieces of 3/4" MDF. Now the holes lining up perfectly would be nice, but certainly not as important as the edges of all the pieces lining up so that the wood laminate pieces have a smooth contour when bent against it.

Taking every piece individually to the drill press is really a non-starter. They are too heavy, and want to tip off the drill press. Plus no way I could remotely handle lifting all 8 pieces together.

So lets assume I attach all 8 pieces plus the pattern together. How do I then, without the drill press, drill those 1-1/2" diameter holes in 6" of MDF accurately? The pattern on top would act as a guide to some degree.

[Phil Gaudio]

It's easy enough to build a temporary auxiliary table for your DP. Here is the project I am currently working on: that table supports the 50 lb slab that will become the seat of a bench.

https://i.postimg.cc/c4tWYPz6/IMG-4171.jpg (https://postimg.cc/F7vwXCNQ)

[Alan Lightstone]

I did just find a 2" diameter hole saw whose description says it can drill up to 6-1/4" depth.
https://www.grainger.com/product/SPYDER-Hole-Saw-2-in-Saw-Dia-38HY29?opr=ILOF

I can't imagine how difficult it would be to remove the 6" plug from the hole saw after each cut.

[Alan Lightstone]

OK. The forms are built. Just waiting for a deeper hole saw to drill holes in the forms for clamps. And yes, perhaps I should have made the middle hollow. It weighs more than I can possibly lit.
517162
Now the $64,000 question.

I was only able to get kiln dried white oak from my supplier. He laughed when I asked for air dried white oak. So clearly not as easy to bend. Oh well, that's what I have. So for technique, looking at the above suggestions I have a question.

No doubt I could make it out of a boatload of bent laminations from the white oak without steaming. But I do have the materials to build a large steam box and just bought the steam generators from Rockler (which can be returned). But will this allow me to use thicker laminations, or is this just a waste of time and money considering the kiln dried white oak that I have?

I was thinking of trying the Michael Fortune technique of steam bending the wood in both directions to make it more pliable, and then doing bent laminations with them, as springback is a total no-go for me. https://www.finewoodworking.com/2007/03/15/free-form-steam-bending-technique

Is this double the work and will the kiln-dried white oak just not steam bend any thicker than just bent laminations? Again, these will be 5-1/2" tall and 1-1/2" thick, so lots of laminate pieces to be bent.

[Steve Demuth]

Alan,

If you're steam bending them before doing a lamination glue up, then a little spring back is not a big problem. A little bit can even be helpful in getting the lams fit onto the form.

As for the steam bending of kiln dried white oak: You should make a test piece 3/16" thick (or maybe 1/4" if you're feeling really adventurous), steam it, and see how it bends. My experience is that at those thicknesses, the difference between air dried and kiln dried isn't all that significant (it certainly is if you go much thicker, but you won't get great laminations from anything thicker anyway). Again, you don't need a perfect, zero spring back bend, because the glue up will take care of any modest spring back quite nicely as long as you use a glue that won't creep on you.

The key to getting near perfect laminations in a situation like yours is to use a glue with adequate open time, and start the clamping at the center of the bend, working out symmetrically. Use lot's of clamps. Roll the laminations down onto each other as you work away from the center. It can be useful, if you can swing it to clamp with cauls that give you a lever point, so you can push each successive caul both down and away to smooth and eliminate any bulges in the lams.

[Alan Lightstone]

Making the steam box is next as I'm finally done with the form.

I've been watching many videos on wood bending. This one from a very talented furniture maker takes things to an amazing extreme. Not my style of furniture, but very impressive work:
https://www.youtube.com/watch?v=eZOY9DBXprI

Another talented Aussie. Lots down there.

[Jim Becker]

Pedulla is one of my favorites. He's referenced the other fellow's work (the guy with the bending fixture) when discussing steam bending, actually.

[Alan Lightstone]

Pedulla is one of my favorites. He's referenced the other fellow's work (the guy with the bending fixture) when discussing steam bending, actually.
The "other fellow" Darren Oates, has unfortunately taken all his content off of Youtube. Pity, as I hadn't downloaded it as I saw no need. It was very instructive, and he produces beautiful furniture.

[Jim Becker]

Strange...he has a large subscriber base. I wonder what happened.

[Alan Lightstone]

So, my next question. I'm building the steam box now (big sucker - 89" long x 12" wide x 7" tall). Considering how large the size of the required steam box is, my question is regarding steam generators.

I purchased two steam generators from Rockler. But should I be using two, versus returning one and just using the single one. I will have pressure relief hole/holes in the box, but I wonder if the far end will get enough steam to help the wood bend. In particular this will be kiln-dried white oak, so that's not in my favor to start with. And if I use two, where to place them.

[Bradley Gray]

I have an Earlex steamer. Works well but only runs for an hour before it's out of water. With two, you could stagger the start. The biggest issue is insulating the box. A thermometer is really helpful. I bend cherry and walnut 1 x1" squares for music stand legs. They need 1 1/2-2 hours. A metal band with end blocks reduced my breakage from 60% to 20%.

[Alan Lightstone]

I have an Earlex steamer. Works well but only runs for an hour before it's out of water. With two, you could stagger the start. The biggest issue is insulating the box. A thermometer is really helpful. I bend cherry and walnut 1 x1" squares for music stand legs. They need 1 1/2-2 hours. A metal band with end blocks reduced my breakage from 60% to 20%.
I thought of staggering the starts, but I'm wondering if a single steamer can provide enough steam for that large a volume box.

[Bradley Gray]

If you're using the Ealex you will have plenty of steam. Go nuts insulting the box and make a steel backer with end blocks. You have to compress the fiber on the inside of the curve and keep the outside fibers from fracturing.

[Alan Lightstone]

If you're using the Ealex you will have plenty of steam. Go nuts insulting the box and make a steel backer with end blocks. You have to compress the fiber on the inside of the curve and keep the outside fibers from fracturing.
How do you insulate the box? It's 3/4" exterior plywood. I'm using the Rockler steam generator.

[Steve Demuth]

How do you insulate the box? It's 3/4" exterior plywood. I'm using the Rockler steam generator.

Whether insulation is worth the trouble depends on where and how you're going to use the box. I don't insulate mine at all, but I mostly use it to steam relatively thin wood (as your project demands), so I don't need terribly long times in the box . If you're going to bend 1.5" wood, you'll need at least an hour in the box as Bradley says, but 10 minutes in live steam will completely "melt" a 1/8" lam strip. Takes somewhat longer if you load a big box up with many such strips, but still, you're talking less than 30 minutes.

Also, understand that if you insulate a plywood steam box well on the outside, you are going to break down the plywood over multiple heating cycles. You're basically guaranteeing that the plywood will be reaching steam temperature almost as quickly as your bending wood does, and that it's going through extreme moisture content cycles. Even exterior grade off the shelf plywoood is not designed for repeated cycles of 200oF temperature cycles.

My advice - do a test run on one or two strips of wood of the thickness your target material is. Bring the box up to temperature, put your wood in, and check after 15 minutes, then again after 20+. You'll likely find strips of the sort you're doing are in great shape to bend.

If you do find yourself in cycles where the Earlex is running low, keep a thermos of very hot (boiling when poured) water at hand. You can top off the Earlex when it gets low, and recovery time for steam will be very low. If you feel some insulation would be of help, you can probably find a cheap thermal blanket at a used household good store, or pickup up a thinsulate camping blanket you can wrap around your box.

[Alan Lightstone]

Making 85" thin resaw cuts from 8/4 white oak is not the easiest thing in the world to say the lease. Spending lots of time moving and readjusting the roller thingies to support the wood before and after the bandsaw, jointer, wide-belt sander.

I haven't bent any wood yet, but getting ready to. Building the hardware portion of the jig now.

I have seen a number of people (including Marc Spagnuolo) soaking the wood beforehand overnight to help with the bending process. I have no idea what I can use to soak 6" wide by 85" long blanks. Not going to build a trough. And don't see any PVC pipe that large (can't imagine what that would weigh). Is the soaking helpful? Necessary? Again, this will be kiln-dried white oak, so I'm behind the eight-ball there to start.

I just got the thermometer to measure the temperature in the steam box. That will show me if I need one or two of the Rockler (Earlex) steam generators.

[Bradley Gray]

How do you insulate the box?

I tried blue foam insulation first but it can't take the heat - rigid fiberglass worked fine. You may be able to overcome the heat loss with 2 steam generators, and you are bending a lot thinner stock than me. for 1" stock to soften, I need the insulation. That, and a steel backing strap for the bend made it all work for me. Your thermometer will tell a lot - you really need 210-215 F and insulation will help you get there. Start timing when you reach that point. I have been using the same ext. ply insulated box for about 5 years and the plywood is still OK. I'm sure I will need a new box at some point but replacing that is less of a problem for me than having a higher failure rate, as nice straight grained walnut and cherry are more dear than ext. ply.

[Alan Lightstone]

Hopefully the second steam generator can overcome the heat loss. I'll be able to check it in a couple of days.

I'm kicking myself for giving away the blowup kiddy pool we had for our granddaughter. That would have fit the wood for soaking. But does soaking help? People mention it with fabric softener added.

[Maurice Mcmurry]

Soaking helps. It can lead to discoloration. I have various pipes with end caps for soaking, including a piece of a plastic culvert. Watlow heaters are also very helpful. Very flat grained plain sawn lumber is easiest to bend. I made a few parts that only worked when green, plain sawn wood was used. Here is a snip about Watlow heaters.

Watlow silicone rubber heaters are rugged, yet thin, lightweight and flexible. Use of these heaters is limited only by the imagination. Heat can be put where it is needed. These silicone heaters improve heat transfer, speed warm ups, and decrease wattage requirements. Fiberglass-reinforced silicone rubber gives the heater dimensional stability without sacrificing flexibility. Because very little material separates the element from the part, heat transfer is rapid and efficient. The silicone rubber heaters are constructed with a wire-wound element or an etched foil element. The heater construction creates a very thin heater allowing it to fit applications where space is limited.
Silicone Rubber Heaters | Watlow (https://www.watlow.com/Products/Heaters/Gas-Delivery-and-Exhaust-Heaters/Flexible-Silicone-Rubber-Heaters)

[Alan Lightstone]

My game plan is to do whatever I can to make it possible to bend thicker laminations. No way I'll be able to bend 1.5" kiln-dried white oak, but also not looking forward to the prospect of having to laminate 24 strips of 1/16" either. 1/4" would be great. 1/2" (not-likely I think) would be awesome. So if soaking can get me to 1/4", that's real progress.

[Maurice Mcmurry]

1/4 inch should be very doable. Banjo rims are done with 1/4" and 5/16" and are bent around a 9 1/2 inch diameter. I agree with the others that the furniture in the first image must be made from veneers.

517779

[Steve Demuth]

My game plan is to do whatever I can to make it possible to bend thicker laminations. No way I'll be able to bend 1.5" kiln-dried white oak, but also not looking forward to the prospect of having to laminate 24 strips of 1/16" either. 1/4" would be great. 1/2" (not-likely I think) would be awesome. So if soaking can get me to 1/4", that's real progress.

You could do it without any steam if you were cutting the lams down to 1/16" - simply bending them around the forms with a backstrap to control them would work. 1/8" with kiln dried, you likely need the steam for the curves you're after - but it's just as much to make sure that the lams set in something approximating their final shape. Beyond 1/8" with quarter sawn white oak, you're going to want them steamed up to temp and to use a backstrap as Bradley says, to prevent tearing on the outer radii (but also note, that since you're laminating them with glue, other than needing one really nice one for the outer bit, a small bit of tearing in the tight bends isn't going to compromise your build in any way, structurally or visually.

For a clue as to how tight you can bend steamed 1/8" lams, look at the narrow end of this basket. This is, to be sure, air dried red elm, so not as prone to breakage as your kiln dried oak, but then, that's a tight bend, too.
517780517781517782

(Just to brag on my beloved, since I'm here - the yarn in the basket is all hand spun, hand dyed with natural dyes, wool from our flock, and the hanks of fiber are stricts of unspun linen she processed from flax that she raised).

[Kevin Jenness]

Wood arrived yesterday. Busy making the forms now. Using 3/4"MDF. OMG the weight of 8 of these on top of each other to accommodate the 5-1/2" wide wood that needs to be bent will require help, to say the lease.

I made the curved portion of the form out of 1/2" MDF using a Carter Accuright circle cutting jig that I had on my bandsaw. What a PIA. Shoulda just made a trammel with my router. Probably would have come out a little smoother too, though mine will work. Killing me the waste of MDF when I'm done. Not that I ever use MDF, but still taking a ton to make the bending form.

I couldn't lift or transport 4x8 MDF sheets, so I'm going to make this out of 2x4 MDF sheets and domino them together.

Now I need to build a larger steaming box than I have, and probably get a larger steam generator. Lots and lots of work to do. Any good suggestions for steam generators. Somewhere I read that wallpaper steamers, whatever they are, are good.

You can steam wood in a bag instead of a box and fold it up after use. https://www.youtube.com/watch?v=50uXPPt8-VI I have used a couple of Earlex steam generators and a 30 gallon bag successfully, you might get by with one and a blanket for heat retention. For larger quantities of steam the propane burner and gas can shown in the video are pretty common.

Have you done a test bend with dry strips to ascertain the thickness needed for lamination w/o steaming? The whole process you are planning is very laborious compared to straight lamination or segmenting the curve. I have never done that, and no doubt it will work but I wonder how flat across the laminations will be after steaming/drying and how tight your finished gluelines will be. Seems like some testing would be a good idea before slicing up all that material.

A small furniture shop near here has done a fair amount of steambending for thick chair parts using a ratchet winch and bending straps with endstops, milling the parts to final size/shape after bending. That would be my preference on aesthetic grounds.

[Alan Lightstone]

Well, the verdict is in on using a single Earlex steamer. The highest temperature it could get to was 174 degrees F after one hour. So I installed a second Earlex which I had, and that gets up to 212 degrees F in about 30 minutes.

Several areas of the box are leaking some water. Can caulk stand up to that heat, or is there something else I can use to seal it from the outside without taking things apart?

I may definitely wind up just using dry bent lamination, but was hoping that by steaming the laminations first and pre-bending them, then later putting them back on the form for a more normal bent lamination I could use thicker laminations. Clearly much, much more work this way. May not be worth it, but well, well down the path.

No question in retrospect that segmenting the curve would have been far easier, though I really don't know if I could produce identical pieces/curves that way. No CNC machine here.

[Steve Demuth]

Steamboxes always leak water. If you want to control the leaks, I would assemble the box with a temperature rated caulk or roofing compound in the joints (and, I predict, it'll still leak some), set the box at a slight angle so the water all runs to one end, and make the water come out through a leak (hole) of your choosing on the downhill end.

Your plan to steam bend, then glue up in a second step is absolutely sound. I've done it many times.and there is no reason for you not to succeed. You should have no difficulty getting 1/4" laminations to work this way, although my personal preference is to use slightly thinner lams than that - usually 5/16" or 1/8" - just because they soften up so much faster in the steam bath. On the other hand, they also cool very quickly, so you need to be quick and efficient in getting them from steamer to form. 1/4" will give you noticeably more leeway in that regard.

[Bradley Gray]

My boxes have a drain hole.

[Alan Lightstone]

My boxes have a drain hole.
Mine does too, as well as a hole on the door side to allow steam to escape. It's the wood joints along its length that have some areas that are leaking.

That being said, I really will need for this to move outside to use. I don't think it's a great idea to be releasing steam in my workshop, rusting all my machines.

[Maurice Mcmurry]

I was impressed by the bag technique shown in the video. He had some serious BTUs under his steam generators.

[Alan Lightstone]

So lots of progress, but a few more questions and a perplexing situation.

I finished building the form, and attached a winch. That took a few extra tries as the wood is so long that I needed more space than I had to allow the cable from the winch to reach the wood pre-bending, yet still be able to tighten down on the form. It turns out that wasn't so easy.
517987517988
So I wanted to make a few packers to sit outside the white oak laminate strips per the online suggestions. I grabbed a piece of spalted maple that I had sitting around for years collecting dust, resawed it to about .15 inches and steamed it for an hour. It's unbelievable how fast those Earlex steam generators go through water. And filling them when hot is no great task either. I wore gloves when taking the board out but chose poorly. Got a small 1st degree burn on one finger. New gloves ordered, lesson learned.

It bent quite easily in the mold (although this is totally a two person job), and I left it in the mold for about 8 hours.

I took it out this evening, and this is what it looked like:
517989
I really don't know what to make of this. The board does not appear to have any cracks, but sprung back massively. Is this just the wood I chose for that? I really didn't want to waste white oak on these packers if not necessary.

Really, really wondering if just making more bent laminations instead of steam bending would be far easier. But I'm seriously down the rabbit hole now, so I'll work on this for a bit.

[Kevin Jenness]

No question in retrospect that segmenting the curve would have been far easier, though I really don't know if I could produce identical pieces/curves that way. No CNC machine here.

I don't see why not. Join the segments, bandsaw the curves and clean up with a segmented flush trim shaper cutter and/or spindle sander. A compass plane could be used with care at the joints' change in grain direction.

[Steve Demuth]

A couple of things:

First, I assume the maple was kiln dried, which makes it harder, as you know. Harder in this case means more springback, because the lignin is already set from the kiln. So that's part of it.

But, second, I suspect you've got a cooling problem. One challenge with steam bending thin strips is that they lose temperature very rapidly when you pull them out of the steam chamber, and with a large piece going in an elaborate bending setup, can easily below the resetting temperature of the lignin in the wood by the time you actually get them bent. When bending thin wood for glue lams, as you are, I would not even use your bending apparatus for any lams 1/8" or less in thickness - I'd just slip them out of the steam, lay them over a curved form of the right radius, and immediatel weigh them down, or clamp them at the ends. You're not trying at this stage with thin lams to get a perfect conformance to the final shape - you'll do that in your big form when you glue them up - but rather, get the lam approximately in shape, and you want to capture the bend when the wood is hot through - especially since you've got kiln dried, recalcitrant wood to go with.

You're elaborate mechanism will still be very useful on the glue up, by the way. It's not a waste. And it'll work great for steam bending too, if you decide to go with thick pieces that hold their heat.

Edited to add: The spalted maple probably didn't help either. Spalting is of course the beginning of wood breakdown. Wood that is losing it's structure doesn't bend as well, or hold a bend as well. So, depending on how far along it was, that could have contributed to your issue.

[Kevin Jenness]

A couple of things:


But, second, I suspect you've got a cooling problem. One challenge with steam bending thin strips is that they lose temperature very rapidly when you pull them out of the steam chamber, and with a large piece going in an elaborate bending setup, can easily below the resetting temperature of the lignin in the wood by the time you actually get them bent. When bending thin wood for glue lams, as you are,

This is said to be one of the benefits of steaming in a bag. The blanks can be heated on the form and bent in the bag while the steam is still live.

[Steve Demuth]

This is said to be one of the benefits of steaming in a bag. The blanks can be heated on the form and bent in the bag while the steam is still live.

Yes, for sure. I have steamed and bent wood in bags. Works great for some of the bentwood basket-like projects I do, particularly as a final step to set the final shape. On the yarn basket I posted pictures of above, for example, I bent the wood out of a steam chamber into rough curves, then clamped the wood into final form, which put a lot of tension back into the strips. Then I steamed the whole thing again in a bag until the wood was plastic, held it at that state for 10 minutes and then cooled it, setting the pieces in the shapes I needed. I then dried, sanded and finished the pieces, before glueing the "neck" of the piece, and riveting the rest for final assembly.

I've also seen boatbuilders use long, thin bags to steam pieces that need a slight bend, because it allows them to use the actual boat as the form for clamping the pieces to final form.

[Alan Lightstone]

Steam bending in a bag is unfortunately a non-starter for me, as no way I want to send that much steam into my workshop and rust everything. That would be outside, and keeping an MDF form outside and somehow making a workbench - this just won't work.

So I think bent lamination it is. This video https://www.youtube.com/watch?v=B4zJsDCfpXs shows a pretty similar sized bend in wood. I'm not planning on using complementary outside forms for clamping, just my jig. But the size of the curved legs on his table are very close to the size of my chair backs.

But here's my next question. At 11:48 of the video he passes the curved, laminated, glued piece through the table saw to get to final width. That just strikes me as incredibly dangerous. Am I wrong? Is there a better, safer way of doing that? My wide-belt sander is not quite wide enough (yup, shoulda gotten a bigger one). Should I try to wide a local cabinet shop that would do that for me? Or is the table saw method safe?

[Jim Becker]

I've seen many makers put curved laminations through the table saw to bring to final width after jointing one edge. Yes, it's a little different operation but with the blade height set carefully and assuming the workpiece isn't really narrow, it's doable with "reasonable" safety. Having a helper on the outfeed side to help insure the workpiece doesn't tip can be a good idea. If the amount of width variation is minimal, after jointing one edge, it's conceivable to dress the other edge with a thicknesser if you can work out proper material support on both infeed and outfeed sides of the machine.

[Alan Lightstone]

I've seen many makers put curved laminations through the table saw to bring to final width after jointing one edge. Yes, it's a little different operation but with the blade height set carefully and assuming the workpiece isn't really narrow, it's doable with "reasonable" safety. Having a helper on the outfeed side to help insure the workpiece doesn't tip can be a good idea. If the amount of width variation is minimal, after jointing one edge, it's conceivable to dress the other edge with a thicknesser if you can work out proper material support on both infeed and outfeed sides of the machine.
What is the proper blade height?

I don't think my planer can deal with a piece of that width.

[Steve Demuth]

Again, I would emphasize that you can get the advantages of pre-bending with steam without a bag. Just take the strips out of the steam, and immediately wrap them around the form and clamp the ends. With 1/8" lams, the "immediately" part is critical because they cool so fast. Bags work great for some situations, but aren't necessary for what you're doing (and you definitely can't steam MDF forms anyway).

[Kevin Jenness]

But here's my next question. At 11:48 of the video he passes the curved, laminated, glued piece through the table saw to get to final width. That just strikes me as incredibly dangerous. Am I wrong? Is there a better, safer way of doing that? My wide-belt sander is not quite wide enough (yup, shoulda gotten a bigger one). Should I try to wide a local cabinet shop that would do that for me? Or is the table saw method safe?

Yes, it is safe enough and commonly done after first jointing one edge. Be sure to keep the work tangent to the table at the front of the blade.

With a continuous radius curve you can set up a block on the outfeed side to support the piece as you feed it through. If the curve terminates in straight sections that won't work and you may want a helper. When you are most of the way through the cut the leading end will come back toward you and can be grasped to help finish the cut.

The blade need be only high enough to cut through the material. If you aren't comfortable using a tablesaw a handsaw or jigsaw will work or use the jointer to work to a line on the second edge. A planer or widebelt will have trouble feeding the lumpy edge evenly. Another alternative is to use a router to trim the layup flush to your form.

[Steve Demuth]

What is the proper blade height?

I don't think my planer can deal with a piece of that width.

Personally, I would not try to cut the U-shaped bends you've got on a table saw. Jim is probably right that with some help it can be done with reasonable safety, but it's not going to be easy, I suspect you'll end up with deep saw marks and burns. Just too much torque wanting to rotate the piece against the blade on a piece like that.

I have successfully jointed pieces like that. The saving grace with the jointer being that I could build a big enough "table" using various roller stands and the like to fully support the piece all the way through its rotation through the jointer. (You will have to rough the second edge down to parallel with the first, using a belt sander or hand plane, of course, since you can't thickness plane on a jointer).

If you don't want to, or can't, do that that, though, I would go after the edges with a handheld belt sander, using your bending form as a guide. That is, I'd clamp the piece back in the form with the line I want to achieve even with the top of the form all the way round, then carefully sand down to the form using the belt sander. Or do the same, but hog off the edge using a roughing plane, before switching to something more refined. When you've got one edge flat, flip it in the form, using blocks underneath the flattened edge to get a uniform thickness "in the form" and take off the top again.

[Jim Becker]

Alan, this new video by Keith Johnson just hit and it might be helpful to add some additional comfort to your project...


https://www.youtube.com/watch?v=bsC9UXoSLjo

[Alan Lightstone]

Personally, I would not try to cut the U-shaped bends you've got on a table saw. Jim is probably right that with some help it can be done with reasonable safety, but it's not going to be easy, I suspect you'll end up with deep saw marks and burns. Just too much torque wanting to rotate the piece against the blade on a piece like that.

I have successfully jointed pieces like that. The saving grace with the jointer being that I could build a big enough "table" using various roller stands and the like to fully support the piece all the way through its rotation through the jointer. (You will have to rough the second edge down to parallel with the first, using a belt sander or hand plane, of course, since you can't thickness plane on a jointer).

If you don't want to, or can't, do that that, though, I would go after the edges with a handheld belt sander, using your bending form as a guide. That is, I'd clamp the piece back in the form with the line I want to achieve even with the top of the form all the way round, then carefully sand down to the form using the belt sander. Or do the same, but hog off the edge using a roughing plane, before switching to something more refined. When you've got one edge flat, flip it in the form, using blocks underneath the flattened edge to get a uniform thickness "in the form" and take off the top again.
I like that idea a lot, Steve. Sounds so much safer. Plus this would easily allow me to ease the edges. Gently, considering the veneer.

With my jointer I could do one edge pretty easily. It would be getting the opposite edge parallel that would be the issue.

[Alan Lightstone]

Alan, this new video by Keith Johnson just hit and it might be helpful to add some additional comfort to your project...


https://www.youtube.com/watch?v=bsC9UXoSLjo
Pretty impressive, Jim. Thanks for posting this.

[John C Cox]

I've done over 1/4" thick stock, 6" wide for guitar sides. It's not hard once you get the hang of it.
Tips:
1. Heat is just as important as wet. Wet can result in delamination and discoloring. A hot pipe and a spritz bottle for water can accomplish magic.

2. Get some really good, waterproof hot gloves.

3. Wood grain matters. The selection of your wood for the straightest, clearest stock possible will massively improve your chances of success. Knots, runout, and shifty grain = cracked bends. Back in Ye Olde Days, they would actually split out stock to be bent with wedges and froes rather than sawing it out. That ensured the grain was as straight and true as possible. I will attest that this greatly improves bending. I know it's not always feasible. Not everybody has access to split stock. I've bent quite a bit of stuff sawn out of boards and it usually works, but sometimes it simply doesn't because of the runout and grain in the wood.

[Maurice Mcmurry]

I've done over 1/4" thick stock, 6" wide for guitar sides.

That must have been a massive guitar. My guitar sides are seldom thicker than .090" and never wider than 4 7/8"

[Alan Lightstone]

So, been busy working away on this (when I have time, which lately has been not enough). Ran into some interesting issues.

1.) I was getting very poor resaw pieces. Big variation in thickness. Looking more closely, my tall bandsaw fence really was making things worse in a way. My resaw fence had the same tall height past the blade until the end of the table.
518743
Made attaching it easier to the factory fence, but I clearly was running into the issue of the wood's internal tension pushing against the fence after the cut and moving the wood. So I cut off the fence about 2" or so after the blade, and that seemed to help. Thanks, Derek for that tip (and others).

2.) Processing 8' long 8/4 white oak is way too heavy for me. Really a bad idea. I was worried about more waste using 5/4 or 4/4 boards. But after several resaw cuts are made, the 8/4 board becomes a much lighter, thinner, easier to handle board. Really will use that approach in the future.

3.) I think, looking at the quality of the resawed boards as they come off the bandsaw, that my blade must be dull. I'm able to fix that with MULTIPLE runs through my wide-belt sander on a sled I built, but that also really reduces yield. I've had to make about .25" resaw cuts, and then sand them down with multiple runs through the wide-belt sander to get to a final thickness of about 0.15". Really wasteful of wood. I chose 0.15" thickness of the laminations as it's an even multiple of 1.5" final thickness, and it seemed I needed to go that thin to get smooth boards off my lousy resaw cuts (and should be able to bend, I hope, considering the kiln-dried white oak I am using.)

3.) I did get double-sized tape to work well keeping the wood on the sanding sled. I remove it with a squirt of denatured alcohol. Amazingly, I called the company's tech support and asked which solvent to use, and they had no idea. Eventually he through out mineral spirits, which does nothing, BTW.

4.) Steaming the pieces (I'm trying a few now to test) is annoying, and it is a massive rush to get them in the bending jig, but I can do two at a time.

5.) Now, my biggest issue. When on the bending jig, I tried for the first time two pieces at a time. But there is a big gap between pieces, despite tightening down the clamps and tightening the winch. How can I fix this? I have a piece of metal going all the way around the jig, in theory pushing the wood towards the form, but those gaps between pieces (which are sanded very smooth), are preventing this from being built.
518740518741518742

Now when that Australian woodworker was using a similar setup, he made a packer (I think that's what he called it) of multiple pieces of veneer and had that inside the metal sheet to even out the pressure. I didn't use that. Is that necessary? Is that what I need? It looks like I need more pressure on the wood, but I don't know how I can do that. Physically, I don't have the strength, for sure, and mechanical advantage doesn't seem to be enough here.

[Kevin Jenness]

Have you considered using air pressure? A section or two of firehose and a caul trapped between a two piece form will supply higher and evenly distributed pressure. Short of that, use cauls and more clamps. Drill more holes, offset from the first ones, and use two clamps in each hole, one up and one down. Use bending plywood and cross-blocks to spread out the pressure.

[Mike King]

If you are getting gaps, then your laminates are likely too thick for the radius of bend you are attempting to get. What is the width of the laminates you are using? Why not use a planer as opposed to the wide belt sander to thickness the laminates?

[Alan Lightstone]

If you are getting gaps, then your laminates are likely too thick for the radius of bend you are attempting to get. What is the width of the laminates you are using? Why not use a planer as opposed to the wide belt sander to thickness the laminates?
The laminates are 0.15" thick. They bend without breaking, fortunately.
I can get very consistent thicknesses from my wide-belt sander. My jointer is a little off side to side (long story). I've also heard that running very thin laminates through a planer can sometimes end in quite the destructive scene.

I'm thinking of using either some wood laminates inside the bending steel to provide a less flexible backing for the bend, or using wide ratchet straps as I've seen some people do. I hate the concept of having to rebuild that jig again.

I've seen 4" ratchet straps, but can't seem to find 6" wide ones.

[Mike King]

The way I was taught bent lamination (by Michael Fortune and Adrian Ferrazzutti) was to thickness the laminates on a planer. Michael uses a lunchbox planer to thickness his laminates; he thinks that the rubber feed rollers provide an optimal amount of pressure for thin laminates. If you are using a larger planer and the laminates are as thin as you state, then you can use an auxiliary bed for the planer: it's simply a piece of melamine with a cleat on it that is less than the width of the planer bed. The cleat goes down on the infeed side of the planer bed and holds the auxiliary bed in place -- wax it, and you have transformed your planer into a thin laminate thicknesser.

The issue isn't whether the laminates bend without breaking; it is whether they bend sufficiently to follow the shape of the form. That's not happening at the moment. Looking at your form and clamping process, you do not have a sufficient number of clamps to apply pressure throughout the entire assembly. Pressure moves from the clamp at a 45 degree angle. With the minimal number of laminates you have and the spring steel cover, you have big gaps where no pressure is being applied to the laminates.

One thing you could try is to use strips of ⅛" thickness hardboard as the cauls on top of your laminates. You'll need a fair number of them to spread the clamping pressure to more of the laminates, say 6 or more. Cover the one against the laminates with packing tape to keep it from sticking to the laminates.

A firehose blown up pneumatically with an outer form could also work. But somehow you need to distribute the clamping pressure over more of the form and laminates.

Mike

[Maurice Mcmurry]

All of my fixtures have as many spots for clamps that the space allows. This fixture looks excellent but there is room for more clamps. For something this wide I make the form flippable and put two clamps in each spot, top and bottom, both of which press on a stiff caul. Commercial set ups for this type of press are massive iron forms with hydraulics.

518868
518871
circular laminating press snip from the web

[Alan Lightstone]

Thanks Maurice, and Mike. Not sure I understand the firehose, though.

Another thought I'm having and am planning on trying is with two 3" ratchet straps. Have to figure out how to use them with this, connection points, etc...

I agree and will have to figure out using some 1/8" wood or hardboard as cauls on top of the laminates. I think this is also a large part of the problem. I didn't realize that hardboard could bend. Never use it or thought about it. The thin metal banding I am using (the thickness was suggested by the metal supplier) is also an issue. Thicker might not bend, but this thin it is not transmitting the force evenly.

Still wondering if not pre-bending the laminates with steam will make gluing up the laminates FAR easier to glue up as opposed to having to spread glue on pre-bent laminates.

I could run my laminates through my planer on the sled I'm using on the wide-belt sander. Outside of saving time and steps, is there some other advantage to that? The wide-belt provides finished laminates that are quite smooth and consistent in thickness. The planer would leave some blade marks. I could see running through the planer once and then the wide-belt, but still not sure of the advantage to using a planer for that.

[Jerry Bruette]

I haven't been following the whole thread but checking in once in a while. It's a pretty interesting project you've taken on.

You could gain some mechanical advantage by going with multiple reaving with your pulleys. Put double pulleys where you have the singles now, then mount singles closer to the winch. Run the cable from the winch through one side of the double sheave, then through to the single sheave and back through the other side of the double and out to your attachment point on the metal. This should reduce the amount of force you need to turn the crank on the winch.

[Mike King]

There are two advantages to using a planer as opposed to a sander to thickness the laminates. The first is speed. The second is that sanding embeds grit from the belt in the wood and that grit will affect any edge tool you use thereafter on the bent lamination.

Pre-bending the laminates isn't going to help you with the issue which is distributing the pressure over the laminates when they are glued up -- you still need to think about the pressure lines off of the clamps.

I've never used firehose myself, but have heard of using it to provide pressure. Here's a link that gives you an idea of how this works:

https://www.instagram.com/p/BBTT9F9oK6U/?utm_source=ig_web_copy_link

There's clearly some issues with this approach: a) there's a lot of pressure involved, and a catastrophic failure can be dangerous, b) I don't know of any sources of information on how to do this -- maybe you can find it, but my googling wasn't promising, and c) the width of the fire hose may not conform to the width of your assembly.

The ratchet straps may be your best solution. However, I suspect that you will still need to clamp the long straights as the ratchet strap is unlikely to apply sufficient pressure to those elements.

One thing that hasn't been mentioned yet is that you need to use an odd number of laminates. This puts the pressure of slip once the assembly is dry in the interior of a laminate rather than on a glue line.

Mike

[William Hodge]

I agree, 2" truck straps work great. Pad the laminations with some bending plywood.

The note above warning about sanding wood to be machined is right on. Any sandpaper touching wood will dull your cutters.

Using clamps to bend wood is so hard to do, it's really awkward and next to impossible to spread the force from such a small clamp foot print.

[Mike King]

Alan, I just went back and reviewed some material I have from Michael Fortune. Take a look at how many clamps he has to ensure that the pressure is distributed through the laminate assembly. Also, notice that he has an interesting caul assembly: a strip of bending plywood with blocks, then some padding (likely hardboard or bending plywood), then the laminates.

518902

He does have a reference to using a pressure hose, but warns that it can be very dangerous...

Mike

[Alan Lightstone]

There are two advantages to using a planer as opposed to a sander to thickness the laminates. The first is speed. The second is that sanding embeds grit from the belt in the wood and that grit will affect any edge tool you use thereafter on the bent lamination.

Pre-bending the laminates isn't going to help you with the issue which is distributing the pressure over the laminates when they are glued up -- you still need to think about the pressure lines off of the clamps.

I've never used firehose myself, but have heard of using it to provide pressure. Here's a link that gives you an idea of how this works:

https://www.instagram.com/p/BBTT9F9oK6U/?utm_source=ig_web_copy_link

There's clearly some issues with this approach: a) there's a lot of pressure involved, and a catastrophic failure can be dangerous, b) I don't know of any sources of information on how to do this -- maybe you can find it, but my googling wasn't promising, and c) the width of the fire hose may not conform to the width of your assembly.

The ratchet straps may be your best solution. However, I suspect that you will still need to clamp the long straights as the ratchet strap is unlikely to apply sufficient pressure to those elements.

One thing that hasn't been mentioned yet is that you need to use an odd number of laminates. This puts the pressure of slip once the assembly is dry in the interior of a laminate rather than on a glue line.

Mike
I'm speechless after seeing that firehose picture. :eek:

[Alan Lightstone]

The ratchet straps may be your best solution. However, I suspect that you will still need to clamp the long straights as the ratchet strap is unlikely to apply sufficient pressure to those elements.

One thing that hasn't been mentioned yet is that you need to use an odd number of laminates. This puts the pressure of slip once the assembly is dry in the interior of a laminate rather than on a glue line.

Mike[/COLOR]
I agree about the long straights at both ends of the curve. I think that will need clamping and may be the big issue. On the other hand, adding a big piece of wood outside the ratchet straps and clamps I would think would distribute that pressure well.

Been also going through my brain about how this could be done with vacuum bagging, but can't quite get my head around that approach.

[Mike King]

For vacuum bagging, you place the laminates in a bag without a platen (with breather mesh to allow the air to evacuate), draw the vacuum, then bend the assembly in the bag around your form. You need to clamp the bagged assembly around the form. It's best to have a polyurethane bag as opposed to a vinyl bag as they are more flexible. Adrian Ferrazzutti showed us this technique for another student in the workshop I did with him at Anderson Ranch. It was cool.

Personally, I think I'd go with the caul approach Michael Fortune used in the slide I posted above...

Mike

[Maurice Mcmurry]

Here are some forms with access to both sides. I will try to dig the round window form out of the shed. It is set up like a BBQ rotisserie.

518947 518948

[Alan Lightstone]

Here are some forms with access to both sides. I will try to dig the round window form out of the shed. It is set up like a BBQ rotisserie.

518947 518948
Interesting, Maurice. I could have done it that way, but I was building to mimic the setup that Australian guy had on YouTube that he removed right as I needed to see it.

I thought that the winch approach would work, and it still might, but straps I think will work better for my particular bend. And I agree, a caul is a must. Plus clamps (probably going to use a couple of parallel clamps with wood boards) to help hold the straight parts.

Would have been tremendously easier just making the making the curved part on the jig, and attaching it to straight sides. Might have to go with that.

[Jim Becker]

You might be able to use shorter "lifting straps" with your existing winch to do the deed in the same way as with just heavy ratchet straps.

[Maurice Mcmurry]

I learned some about the forces required to make bent laminations from the video about the making of Stienway pianos. "Work around from the center outwards" . The bending and clamping is shown right at the begining.


https://youtu.be/jAInt7hIZlU?si=WUKDhI0Opvke4C8A

[Mike King]

The Steinway factory tour was one of the best things I ever did in NYC. It's too bad that they have now eliminated it. My wife and I tried for years to get tickets to the tour on one of our trips to the city, and when we finally were successful, it was a big day! I'd love to go back...

[Warren Lake]

Where do you get lifting straps? I use chains on the machines and they are great as you can lock on any place and adjust lengths and more. Only negative is the weight of a number of 3/8" chains.

Excellent you tube, very interesting how they jigged up for different operations.

[Maurice Mcmurry]

I used boat winches with seatbelt style webbing. Even with two 2 inch straps and a ton of pull on each, clamps and cauls were still required to flatten the layers.

https://sawmillcreek.org/attachment.php?attachmentid=517756&d=1711847387

[Alan Lightstone]

I used boat winches with seatbelt style webbing. Even with two 2 inch straps and a ton of pull on each, clamps and cauls were still required to flatten the layers.

https://sawmillcreek.org/attachment.php?attachmentid=517756&d=1711847387
I'm coming to that realization. This bending stuff - not easy to do.

My next test will be with two 3" straps and clamps. Then I'll make some cauls for the straight parts. If that doesn't work.....

[Jerry Bruette]

Where do you get lifting straps? I use chains on the machines and they are great as you can lock on any place and adjust lengths and more. Only negative is the weight of a number of 3/8" chains.

Excellent you tube, very interesting how they jigged up for different operations.

You can get lifting slings here. https://www.mcmaster.com/products/slings/slings~/

[Alan Lightstone]

OK. Some steps forward, some steps back.

1.) Clearly the cauls on the straight part of the jig work very well. Clearly make the gaps disappear or get very small.

2.) I opened up the Amazon package and took out the 3" wide ratchet straps. OMG are they heavy and the metal ratchet is way too wide. No way I can use two of these at the same time (could use two straps, but don't know a way to keep the ratchets from getting in each other's way. Kinda looks way too heavy duty for the task, but that may just be by virtue of them being 3" wide straps.
519041
Now the strap did do a very good job tightening the laminates against the jig, at least on the curved part. The straight part clearly needs cauls added to the outside and clamped. That's easy.

But trying to use the winch (which is great for pulling the laminates around the jig), at least how I have it attached with eye bolts gets in the way of the ratchet straps.

So kinda stuck thinking of a game plan going forward.

[Mike King]

Go back and look at the caul and padding photo from Michael Fortune I posted. If you make a similar caul, use sufficient flexible padding, and increase the number of clamps, you will be fine. Michael has bent laminated tighter curves than you are attempting.

[Alan Lightstone]

So going to try gluing some laminations together, then putting them in the jig. Clearly I need the winch, as I don't have the physical strength to pull the parts together by myself. I'm narrowing down some ratchet straps to help. Clearly this will be a test piece, with yet more modifications to come.

A quick question. For glue, my choices are Ultra-CAT Powdered Pre-Cat Resin, or West Systems Epoxy. I clearly need a very rigid joint to avoid any springback. The pieces will live outdoors in Florida, so heat, humidity, moisture etc... will be a part of their life. I was thinking that the high clamping pressure might make epoxy a great choice, but may be mistaken there for this application.

Any suggestions as to which of these will be better to use?

[Jim Becker]

I would use epoxy resin both for the long working time and the bond. West Systems, Total Boat, etc.

[Michael Burnside]

I agree with Jim. However when you say "I was thinking that the high clamping pressure might make epoxy a great choice, but may be mistaken there for this application.", remember that epoxy doesn't need a lot of clamp pressure to work. In fact, too much and it is not a good idea. Epoxy has gap filling properties which may help you here on minor variances.

[Alan Lightstone]

I agree with Jim. However when you say "I was thinking that the high clamping pressure might make epoxy a great choice, but may be mistaken there for this application.", remember that epoxy doesn't need a lot of clamp pressure to work. In fact, too much and it is not a good idea. Epoxy has gap filling properties which may help you here on minor variances.
Whoops. I typed that backwards. What I meant to say was that I was worried that high clamping pressure would make epoxy a POOR choice. I corrected it above.

[Mike King]

For a really rigid glue line, I'd use Unbound 800. It is ok in exterior applications as long as it isn't continually exposed to water.

[Maurice Mcmurry]

I use West System epoxy thickened to paste consistency with collodial silica. It wets out amazingly well without running all over the place. The exceptional wet out characteristics insure that it will stick strong whether its quite thick or very thin.

[Alan Lightstone]

I use West System epoxy thickened to paste consistency with collodial silica. It wets out amazingly well without running all over the place. The exceptional wet out characteristics insure that it will stick strong whether its quite thick or very thin.
I tried that route yesterday with a couple of layers. Going to take off the clamps today (48 hrs). I'll then see how much springback (if any) I get from the first two inside layers. Strengthwise, I don't think I would be able to glue all 10 laminates together and then bend on the form. So my game plan is to do 2-3 layers at a time, perhaps up to about 4 layers, then winch in place and place clamps.

I also bought some West Systems extra slow hardener to get really long working time.

[Alan Lightstone]

OK. That's it. Waving the white flag. This just isn't working for me. With my setup, physical strength, etc... Steam bending or bent laminations are out for this chair. The pieces have springback, gaps between lamination, uneven height. Just not at all usable and a tremendous waste of wood. Was a somewhat interesting, but mostly frustrating exercise. Spent a good deal of money on winch/straps/mdf/ etc... And have taken up most of my workshop with jigs, etc... for this. Anyone need 2 steam generators and an extra long steam box (sigh...)

So I still need to make the chair (we've already bought the expensive cushions and really need replacement chairs out there. Building them like the originals (solid wood, but not using teak) is the new plan. A whole new set of questions pops up then.

To refresh everyone who's been so helpful memories, this is what the chair needs to look like:
519366

The most difficult part is the chair back, which is 5-1/2" tall and 1-1/2" thick. I can easily make a pattern of the shape and rough cut it on my bandsaw. But how do I smooth out the profile with a router table when it's that tall? Are there pattern bits that long that can safely be used on a router? No shaper here. I need to make 4 chairs, so will need to make a bunch of identical pieces. And my initial concern about lining up the pieces for gluing still seems difficult (but not impossible.)

So mostly concerns about routing the pieces identically and smoothly. Both the inside and outside curves.

[Kevin Jenness]

Forget pattern routing. Do the joinery and assembly, then saw the parts out as accurately as possible and clean up the faces. In my shop I would use an edge sander.

[Mike King]

I realize you are frustrated, but the best approach to achieve that shape is bent lamination. If you really want to get it to work, it seems like you need some expert instruction. Here's the solution to that:

https://schoolofwoodwork.com/event/2024-parker-designfabrication/

Kelly Parker interned with Michael Fortune and teaches woodworking with him quite frequently. If you were to sign up for that class and take your project there, Kelly can help you figure out the caul, padding, and clamping that will allow you to complete your project.

Mike

[Mike King]

An alternate way to achieve the curve would be to Cooper the bend. But rather than grain following the curve, the grain would be vertical (I don't think it would be a good idea to glue end grain to end grain, especially outdoors). You would have a fair amount of hand work to fair the curves.

This brings up a design question: what is the radius of the curve for your chair back? Have you built a prototype to test the comfort of your design? Generally speaking, a comfortable radius on a seat back is 18+ inches.

[Michael Burnside]

One thought I had was what about building them in layers of say 1 to 1-1/2" tall and stacking them. You could easily build a template, rough cut on the bandsaw and flush trim, then stack them until you reached the final height. I guess after that you could veneer and hide the layering or simply lean into it, which is what I would do.

I personally applaud you for your efforts as I would have waived the flag ages ago. I'm a serious hobbyist and once all the fun is gone and it becomes a job, I've done something wrong...

[Alan Lightstone]

An alternate way to achieve the curve would be to Cooper the bend. But rather than grain following the curve, the grain would be vertical (I don't think it would be a good idea to glue end grain to end grain, especially outdoors). You would have a fair amount of hand work to fair the curves.

This brings up a design question: what is the radius of the curve for your chair back? Have you built a prototype to test the comfort of your design? Generally speaking, a comfortable radius on a seat back is 18+ inches.
A really interesting suggestion, Mike. Out-of-the-box thinking. Don't think I'd like the look, but very interesting.

[Alan Lightstone]

One thought I had was what about building them in layers of say 1 to 1-1/2" tall and stacking them. You could easily build a template, rough cut on the bandsaw and flush trim, then stack them until you reached the final height. I guess after that you could veneer and hide the layering or simply lean into it, which is what I would do.

I personally applaud you for your efforts as I would have waived the flag ages ago. I'm a serious hobbyist and once all the fun is gone and it becomes a job, I've done something wrong...
Probably my approach. I can probably find a long-enough pattern bit to do this in two 2-3/4" tall layers.

Of course that still leaves the problem of gluing those curves together (dominos I would think), and repeatability.

[Michael Burnside]

Probably my approach. I can probably find a long-enough pattern bit to do this in two 2-3/4" tall layers.

Of course that still leaves the problem of gluing those curves together (dominos I would think), and repeatability.

They won't be curves, they will be 30 or 45 degree miters that you will make curves after the glue has dried. I would make a template that is the exact curve of the entire chair back. Do NOT use a Domino unless you are 100% sure you won't expose them when making the curves. Personally I would offset the wood so it's long-grain adhesion that is the strength and just use some CA glue and wood-glue combo to hold the miters while they dry. No dominos needed.

I'm actually working on a similar design for a console that is curved on one end. It's 16" deep so I will be gluing 16 1" curves together. So, misery loves company :D

[Mike King]

They won't be curves, they will be 30 or 45 degree miters that you will make curves after the glue has dried. I would make a template that is the exact curve of the entire chair back. Do NOT use a Domino unless you are 100% sure you won't expose them when making the curves. Personally I would offset the wood so it's long-grain adhesion that is the strength and just use some CA glue and wood-glue combo to hold the miters while they dry. No dominos needed.

I'm actually working on a similar design for a console that is curved on one end. It's 16" deep so I will be gluing 16 1" curves together. So, misery loves company :D

My first curved piece was an A/V console with curved sides that is roughly 20" deep and veneered. I designed and made it with the assistance of Michael Fortune. To make the sides, I kerf-bent ¾" MDF around a form in a vacuum bag, then used the form to veneer the MDF. It was very successful.

519411

[Mike King]

That photo posted upside down. Who knows what is wrong with this forum software...

[Kevin Jenness]

If you feel you need to bricklay and pattern-rout stacked layers in order to get an accurate curve I suggest keeping the thickness down and use 3 or 4 layers as trimming with a long router bit invites chatter. Saw as close to the line as possible. You may get some router tearout near the joints due to grain direction reversal and you will have a busy looking assembly. For exterior use reinforcing the endgrain joints with dominos and epoxy is a good idea.

Bricklaying will give a somewhat stronger result than full height segments. Veneering over the bricklaid assembly will strengthen the back further. You will have to decide if the extra work is worth it. The original makers kept it simple and I would do the same, segmenting the curve in full height pieces with splined and epoxied joints, then sawing and sanding.

The strip laminated approach can work, you need thicker cauls and more clamping force.

[Mike King]

...
The strip laminated approach can work, you need thicker cauls and more clamping force.

And more clamps.

[Kevin Jenness]

And more clamps.

Not necessarily. Pneumatic pressure is an option (Fine Woodworking #63 p. 83, Pneumatic Lamination). If one runs short of clamps threaded rod is an alternative as shown here. https://woodarchivist.com/3725-making-bent-lamination-curves/ In any event the project needs enough force distributed evenly through cauls to pull the laminates together gap-free.

[Alan Lightstone]

I think I would have done far better with air dried white oak, but couldn't get it from my local supplier. That led me down a path that despite tremendous effort, just won't work for me. It's going to be solid wood.

I understand that the joints will be miters. I wrote that badly. If I have to veneer over the outside of the pieces to make it look better, I certainly can do that. Might have to buy a slightly larger vacuum bag, but really don't know if I'll need to go that route.

That's an interesting insight, Kevin, regarding long router bit chatter. I do have a good bandsaw, and a good Hammer oscillating edge sander, so I should be able to sand things reasonably well after gluing, although have to make an extension of sorts to keep the curved piece level when sanding. Sanding the inside of the curve might be a little more challenging.

[Maurice Mcmurry]

Green wood bends the best. For my "tree to finished product" projects I have done the bending with wood that is not even air dried.

[Mike King]

Not necessarily. Pneumatic pressure is an option (Fine Woodworking #63 p. 83, Pneumatic Lamination). If one runs short of clamps threaded rod is an alternative as shown here. https://woodarchivist.com/3725-making-bent-lamination-curves/ In any event the project needs enough force distributed evenly through cauls to pull the laminates together gap-free.

Pneumatic pressure is effectively more clamps. And potentially dangerous, as said well above.

[Mike King]

I think I would have done far better with air dried white oak, but couldn't get it from my local supplier. That led me down a path that despite tremendous effort, just won't work for me. It's going to be solid wood.

I understand that the joints will be miters. I wrote that badly. If I have to veneer over the outside of the pieces to make it look better, I certainly can do that. Might have to buy a slightly larger vacuum bag, but really don't know if I'll need to go that route.

That's an interesting insight, Kevin, regarding long router bit chatter. I do have a good bandsaw, and a good Hammer oscillating edge sander, so I should be able to sand things reasonably well after gluing, although have to make an extension of sorts to keep the curved piece level when sanding. Sanding the inside of the curve might be a little more challenging.

Kiln dried lumber is not an issue with bent lamination. You need 1) laminates that are thin enough to be able to bend around your form, 2) an appropriate glue that has a hard glue line (and likely sufficient working time), and 3) sufficient pressure distributed over the laminates to bend the laminates gap free around your form. I've constructed a number of chairs using both steam bending and bent lamination techniques with kiln dried wood.

I gather that you also tried steam bending. A few questions about that:

1. Did you soak the wood? When I bent 10/4 kiln dried cherry, I soaked the blanks for 7 days before steam bending them.
2. Did you wrap the blanks in plastic when you steamed them? This is a technique I learned from Michael Fortune. Wrapping the blanks in plastic helps to retain moisture in the blank as you steam it and increases heat transfer to the blank.
3. How long did you steam the blanks? I'm at 5400 feet so I can't get my steam box above 197 degrees. As a result, I steamed the 10/4 blanks for 4 hours to ensure that they were very well heated. I've also bent 9/4 kiln dried walnut at 8200 feet -- sufficient time soaking to raise the moisture content of the blank and sufficient time in the box to heat the lignon is needed to be successful in bending.
4. Did you use a strong compression strap? Steamed wood will compress considerably, but if a compression strap doesn't tightly hold the inside of the stock, the blank can separate or crack. The Veritas compression strap and fixtures (from Lee Valley) are very good--they were designed by Michael Fortune, and come with a guide to steam bending that was written by him.
5. Did you put your bent parts on a drying form? Once bent, the part needs to be clamped to a drying form to dry. It takes 1 to two weeks for the part to return to equilibrium moisture if it has been properly steamed. Removing the part from the drying form generally leads to a minimum of excess springback or, for that matter, the part looking nothing like the form.

Mike