So, what is steam bending? It's the process of saturating and heating the polymeric cells of the wood so that it will stretch. Question? Why does oversteaming cause the wood to crack and or break? I don't know. Can you tell me? All enquiries, opinions and offerings will be posted. Anyway, the steaming process is something that is supposed to, from my readings, take a specified amount of time. I submit that, regardless of the thickness and width, the type of wood is also a variable that must be considered. Unfortunately, my first experience was with Teak. Very expensive stuff. I have bent a lot of walnut and oak simply by soaking and clamping for a number of days. The top of the walnut display is an example. The thing is, is that the curves were not that severe and in most cases a whole bunch of glue and a nice fit and some good clamps did the job. The severe turns on the teak bench required a lot of flexibility. I felt it on some of the pieces but not all of them and they broke or cracked.
What would happen if we tried to bend a piece of hard maple. (Something I am priming up for.) It seems to me that, maple, being the denser of the woods, would require more time in the chamber. What about it?
Speaking of the chamber, I built mine out of a section of three wall stove pipe. I made a special cowl to induce the steam into the main pipe only. The thinking was that the other two outer layers would retain the heat better, which is required for a successful steam, as I understand it. The other thing is that you want to have a FLOW of steam through the chamber. I suppose almost anything would do for a chamber. I have seen and read of people using a simple pine tube of 1x stock just simply nailed at the edges. I guess if it flows the steam it works. A leaky chamber is apparently not a problem. My thinking is that a leaky chamber is not a problem, provided you have a reliable and forcful head of steam.
That leads us to the boiler. How do you boil water enough to supply the steam? Currently, I am using a tea kettle, a radiator hose and some fittings. I don't have a good feeling about the boiler end of things. I am using a hot plate or my wood stove to boil the water. I think a more substantive heat source is required and a more high tech tank for the boiling of the water. It seems to me that once you get a good, substantive flow of steam going, you should be able to do the job.
DON'T HESITATE. Be ready for the piece when it comes from the chamber. Practice your setup before you ever even boil the water. How are you going to secure the piece in the mold or where ever you are going to form it? Be ready when the sucker comes out, because it's hot and it won't stay pliable for very long. I would suggest that if you have a complex bend to do, that you secure the most sever parts of the bend first. That's where you really need the pliability. The gradual stuff will pretty much happen on it's own.
But, then, there is the question of gluing. Do you glue it when it is in it's steamed state or let it dry and then glue it? I had to glue 8 strips together and I was wishing I knew the answer to that question. I personally waited a day and let the curved pieces sit in the mold and clamps and went back the next day and glued everything because I was afraid of what the moisture would do to the glue up.
Well, enough for now. If any of you have offerings or questions, please submit them to me via the guestbook or email. We'll work on it together.
Here is something I found in a newsgroup that helped me get started.
Subject: Re: Bending wood
From: gregg@clark.harvard.edu
(Gregg Germain)
Date: 1998/02/04
Message-ID: <34d8bcb0.0@cfanews.harvard.edu>
Newsgroups: rec.woodworking
[Subscribe to rec.woodworking]
[More Headers]
TimOdell (Timothy.Odell@sf.frb.org)
wrote:
: Does anyone know of where
I can get information on ways to bend wood?
Below you will find my FAQ on
Steambending which is posted in this
newsgroup from time to time.
--
This FAQ on bending wood is provided
courtesy of Gregg Germain. Any
comments would be welcome. Comments
should be directed to
gregg@clark.harvard.edu.
I've been in the business of
steambending wood for about 10 years now.
I've built a variety of steamboxes
and tried a number of boiler
systems. What you see written
here is a combintation of reading and
actual experience. Mostly experience.
All of my steam bending has been
with either Oak or Mahogany. I've
never tried any other wood as
I do this work in my
boatbuilding/restoration. So
I cannot comment authoritatively on
bending other woods like cedar,
pine, poplar etc.
And if I haven't actually DONE
it, I will not comment on it. I will
not state anything here that
I have ONLY read out of a book and not
tried.
With that in mind, let's fire up the boiler....
To start with there are several rules of thumb which work quite well.
What you are doing when you are
steaming wood for bending, is
softening the hemicelluloses.
The celluloses are polymers that
behave the same as thermoplastic
resins. [My thanks to John McKenzie
for the last two sentences].
And you need BOTH heat and steam
for this. I realize that some people in
Asia "fire" bend their wood
but invariably, that wood is quite wet -
typically quite green. The Norse
boatbuilders used to get their planks out for
shipbuilding and sink them into
a salt water bog to keep them limber until the
time came to use them.
However, we are not always so
lucky as to get green
wood for our bending and you
can have great success with kiln dried wood.
It's useful if you have the
ability to soak your wood for a few days
so that the moisture content
rises - those Vikings knew what they were
doing.
You need heat and you need moisture.
The primary rule is the one for steam time:
One hour of steaming per inch thickness of wood.
I have found that you can OVERSTEAM
as well as understeam. If you
steam an inch of wood for an
hour, try to bend it, and it cracks,
DO NOT assume that you haven't
steamed it enough. There are several
factors involved which could
explain the result - but we'll get to
those later.
It is smart, however, to have
a piece of stock in the steam box that
is the same thickness as the
piece you wish to bend, and that is
expendable. PREFERABLY a piece
taken from the stock itself. Steam that
with the target piece, and after
the requisite steaming time, take the
test piece out and try to bend
that to the mold. If it snaps, then
give your piece MAYBE 10 minutes
more. But no more.
The wood:
Generally it is best if you can
get green wood. I know that this
makes the cabinetmakers among
us shudder. But the plain fact is that
green wood bends easier than
dried wood. I can take a 6 foot
long, one inch thick piece of
white oak; clamp one end to the bench
and hand bend the piece to the
curvature I need - green wood is THAT
limber. However it won't stay
bent, of course, so I steam it anyways.
In boatbuilding, rot is the
main evil.
For those of us that have to
worry about rot, the act of steaming
green wood removes the tendency
of green wood to rot. So no
worries there - boat ribs are typically
made from steam bent oak and
will not rot in a well cared for boat.
But i've done a lot of steaming
of kiln dried oak and it works fine
too.
One thing you want to try to
avoid in your selection of wood for
bending is grain runout. This
will promote cracking when you bend.
Steamboxes:
It is not necessary - and is
in fact detrimental to the bending
process - to have a steambox
that is absolutely airtight. You WANT
steam to be emanating from the
box. If you don't get a flow through of
steam you will not be able to
bend the wood - it will crack as if you
steamed it for only 5 minutes.
I know - I've created a lot of kindling in this manner.
Steamboxes can come in many shapes
and sizes. You want one big enough
so that you can suspend the
wood off the surface, and get a good flow
of steam around most of the
wood surface. A box made of 2 x 8 pine
boards will work. One suspension
method is to drill a hole through the
sides and run a hardwood dowel
through. The dowel holds your wood up
and minimizes the amount of
wood touching a surface. You don't want
the box to be SO big, however,
such that the amount of steam your rig
generates is too small to fill
up the box. You want a wet, steamy box
BILLOWING steam. So the box
has to be sized to the boiler (or the
boiler sized to the box ;^)
).
I have 2 boxes:
For small stuff like 1 1/2 x
5/8 by 6 foot long oak for ribs, I use a
2 inch diameter piece of PVC.
I have it resting on a 2x4 so that it
won't deform under the heat.
I've also nailed sides to the 2x4 so that the
tube doesn't flatten. For a
boiler i use a whistling tea kettle
with the whistle and top taken
off. A length of radiator hose connects
the kettle to a suitable reduction
on the end of the PVC. For a heat
source i use one of those counter
top electic burners.
Works great.
When I had to steam bend 17 foot
long, 7 inch wide, 3/4 inch thick
mahogany for the new cabin trunk
of my boat, I used a steambox built
with 2 x 12 inch pine. For a
boiler i had a 20 gallon steel boiler.
Heat source was a propane burner
I bought at Ace Hardware Store. This
burner is GREAT because it's
convenient and mobile. It generates
45,000 BTU of heat. It's an
aluminum
bowl on 3 legs with one burner
about 8" in diameter.
Lately, I noticed a 160,000 BTU
propane burner in the West Marine
Catalog for $50. I bought it.
Now I'll be able to generate enough
steam to bend ribs for the Constitution.
Now when I say "one hour of steaming
per one inch of wood" I mean one
hour of SERIOUS steam with NO
interruptions. Therefore you have to
pick a boiler whose capacity
will be sufficient for the steam time you
are looking for. I have used
a 5 gallon UNUSED gasoline can for this
purpose.
NEVER put the wood in the steambox
unless you have full steam and the
box is completely filled. Be
ABSOLUTELY certain that you don't run out
of water BEFORE the necessary
steam time. If you do, and are forced to
add more water give it up...you'll
generate kindling.
One way of maximizing the water
use is to have the box tilted at an
angle so that any condensation
within the box runs BACK towards the
boiler.
Another way is to set up a siphon
system so that the boiler is
constantly being refilled at
the rate at which water is boiling off.
A crude ascii picture of this
follows:
hhhhhhh s Where: b's are the
boiler
h h s s's are the steam outgo
| h | h bbbsbbbb l's are a level
indicator:
| h | l b b metal tube from
side
|wwwwwhw| l bwwwwwwb of boiler
with clear
| h | l b b plastic vertical
tube
| | llllllllb b h's is water
hose from
| | b b auxilliary tank
--------- bbbbbbbb w's are water
levels
Aux tank boiler
As the water in the boiler evaporates,
the siphon brings more water
from the auxiliary tank. the
level gauge is a simple metal tube
extending from the side of the
boiler with an elbow pointing up. Over
the elbow you slip a piece of
clear plastic. This way you can
observethe level of the water
in the boiler. The feeder hose from the
aux tank fits inside the clear
plastic level hose so that you can get
inflow and still see the water
level.
One important point:
If you find you have to add water
to the auxilliary tank, be
sure to add water a LITTLE BIT
AT A TIME. Otherwise the flow of cool
water into the boiler will inhibit
the boiling and you will get an
interruption in steam generation:
not good.
It's also best to begin with
a full aux tank to start with so
that you minimize the need to
add cool water to the aux tank. I like
to leave a little air space
in the boiler when I begin.
Many steam boxes have a door
at one end to allow you to slide in
pieces when you want to - and
take them out when needed. For example,
in ribbing out a boat - something
you'd like to do in a day if you
can, you crank up the boiler
and (when steam is up) you put in your
first piece of wood. 15 minutes
later you put in the second. Fifteen
minutes later the third and
so on. Then, when the first piece is
ready, you yank that out and
bend it. This is all supposing that the
process to bend and install
the rib takes less than 15 minutes. When
the first rib is in, the second
piece of wood is ready..and so on.
This allows you to do a great
deal of work while avoiding
oversteaming.
The door serves another important
function. And the door doesn't have
to be solid either - on my small
steam box i LOOSELY stuff in a rag. I
say loosely because you want
steam to be able to come out of the end
(remember you need steam flowthrough).
The secondary purpose is to
preclude cool air from entering
the steambox underneath the suspended
wood.
Bending:
Assume you have the wood cooking
(it makes a nice smell) and the jig
is ready. Take pains to place
everything so that the operation of
removing a piece from the box
and bending it is a FAST SMOOTH
operation. Time is CRITICAL.
You have only seconds.
When the wood is ready take it
QUICKLY out of the box and bend it.
GET CURVATURE ON THE WOOD!!!!!!!!!!!
As fast as humanly possible. If
inserting the wood on the jig
is complicated, bend it with your
hands (if possible).
On ribs for my boat - where there
is a curve in 2 directions - I
take it out of the box, slip
one end into a brace and bend that end
then bend the other end with
my hands. Try to bend it MORE than the
amount you need in the jig.
But not too much more. Then
slap the wood on the jig.
But I repeat you MUST get curvature
on the wood immediately - like
within the first 5 seconds.
Every second the wood cools it becomes
less flexible.
Length of wood and curvature at the ends:
There is practically NO WAY you
can cut a piece to exact length and
expect to get curvature near
the ends. You simply don't have the
strength and you will be thwarted
by springback.
By the same token, if all you
need is a 3 foot length, and
the wood is greater than, say,
1/4 inch thick, you had better cut the
piece 6 feet long and bend THAT.
You can trim the wood to fit later. I
am assuming the lack of some
sort of hydraulic press in your shop - i
know i don't have one. Cut the
stick overlong remembering that the
shorter the stick the harder
it is to bend.
And if you cut it overlong, you'll
have more curvature near the final
finished end - the last 6 inches
of a 1 inch thick piece of oak will
be dead straight. Depending
upon the curvature you need, you may have
to resort to carving the curvature
out of the end of the wood and
should size it with that in
mind.
Jigs:
When you steam bend apiece of
wood, and clamp it to a shape, you wait
24 hours for it to cool thoroughly.
When you take it off the jig, that
wood will spring back somewhat.
How much depends upon the grain and
the type of wood - it's hard
to say. If your stock has a natural curvature
in the required direction to
start with (I try to take advantage of this
whenever possible), you will
get less springback.
So if you have to get a certain
curvature to the final product, make
your jig with greater curvature.
How much?
Tis is the realm of black magick
and I can't personally give you a
figure. One thing I DO know
is this:
It's infinitely easier to unbend
some wood that was overbent,
than it is to put MORE bend
in a cool piece of wood (assuming you
don't have incredible leverage).
Once caveat: if you are bending
pieces that will be glued together to
form a laminate, be sure that
the jig is the exact shape you need at
glue time - I rarely get much
springback from well bent, glued wood.
There are an infinite variety
of jigs you can build. No matter what
type you choose, you can't go
wrong if you own a clamp making factory
- you can never have too many
clamps. If you are bending wood greater
than 1/2 inch thick you must
see to it that the jig is built extremely
strongly: the amount of stress
on it is quite high.
Quite often people will use a
metal strap along the outside of the
wood as they bend. This helps
to distribute the stesses along the
length of the wood and helps
to prevent cracking. This is especially
true if you get grain runout
at the outside edges.
Well that's all i can think of
now. If i think of more I'll add it to
the FAQ.
What changes occur to wood when
it is permanently bent with the aid of steam? -- MH, Perth, West
Australia
The
main structural component of wood is cellulose, a polymer (plastic) consisting
of long molecular
chains
of sugars. While cellulose is extremely useful and is by far the most common
polymer/plastic in
the
world, it can't be melted because the temperature at which its molecular
chains begin to move
relative
to one another is above the temperature at which those molecular chains
begin to fall apart.
In
short, cellulose decomposes before it melts. Shaping or reshaping cellulose
is very difficult, though
chemical
processes have made it possible to reform cellulose into such materials
as cellophane and
rayon.
The
process you describe, bending wood while heating the wood with steam, takes
advantage of the
fact
that cellulose molecules bind strongly to water molecules and that the
water molecules then
lubricate
the chains so that they can move relative to one another. Water is said
to be a "plasticizer"
for
cellulose. Heat, water, and stress allow the cellulose chains to slide
slowly across one another.
With
enough patience, the wood's internal structure can be changed forever.
When the heat, water,
and
stress are then removed, the wood keeps its new shape.
STEAMBOX AND STEAMBOX REDUX
By David Goodchild
I thought I had a really nifty idea for a steam box.
We have an old (1920) stone house
in Philadelphia and these types of houses are nearly impossible
to insulate. The walls are 18"
thick but of this total dimension, about 14" is the stone. There is a
lathe and plaster wall over
small furring strips attached to the headers and the space between the
lathe and the stone walls is
only about 1 1/2" - 2". Not useful for any kind of insulation.
When we remodeled a few years
ago I came up with the idea of insulating the walls (which sorely
needed it) by gluing rigid foam
insulation to the existing walls, applying more dry wall over the
foam, and then building out
the window mouldings to accomodate the additional 1 1/2" of the
combination had added to the
walls.
It worked wonderfully. The house
is much warmer in the winter and almost feels air conditioned in
the summer the insulation does
so much good.
Well, I had some of the rigid
insulation left over and I thought it would make a really good steam
box. You know, easy to fabricate
with some duct tape, good insulation value to hold in the hot
steam, light and easy to move
around, cheap (zero) and all those other good things.
All went well until the second
time I used it. The first time it was a dream. All the above and more.
The steam stayed hot, the veneers
bent easily and the job went quickly.
But the second time I went to
use it, the "steam-box" had already become a limp and useless
wreck. The steam penetrated
the aluminum facing over the foam, saturated it, and caused it to
lose any minor structural properties
that it ever might have had.
End of phase 1!
Steam Box Redux
This was not the only steam box
attempt that went awry. I had previously tried to use some 12"
diameter smoke pipe but this
also did not work well. I learned one thing from this attempt though.
If you are going to make a steam
box, make one that is not TOO MUCH BIGGER than the material
that you are going to steam
within it! That was the problem with the 12" smoke pipe; there was just
too much internal volume and
the steam cooled much too quickly.
I did make a better steam box
after these failures however. It was almost as simple, not as light,
but worked like a charm and
has been working for a couple of years now. It is simplicity itself and
the sketches show how.
I took two 12' 2 X 4's and cut
a 1/4" deep by 1" wide rabbet top and bottom. Then I cut four pieces
of 1/4 plywood to the size (width)
that I wanted to achieve for the steam box. I screwed these ply
pieces down into the rabbetts
in the 2 x 4's. I put a little patch over the joint where the two pieces
of
plywood met. Then I made a plug
for one end and screwed this into place. Then I made a plug for
the other end which was a loose
fit and had a handle on it. This would be the removable one. A
loose fit, because I didn't
want steam pressure blowing it out all over the neighborhood.
For a boiler I used a turkey roaster!
I cut a hole in the top and stuck
in a piece of 3" smoke pipe and with a couple of elbows I connected
it to the bottom of the steam
box through another hole. I sealed everything up with some caulking.
For a heat source I used our gas stove with the steam box stuck out the window.
Worked great.
Of course, if you have some rigid
foam insulation lying around, there is nothing to prevent you
from gluing it over the plywood
and thereby keeping the interior even hotter!
There are only two safety considerations
to observe when building and using the steam box
described above.
1. Proclaim your innocence until
your dying day as to what happened to your wife's turkey roaster.
Blame it on the turkey roaster
gang that has been pillaging the neighborhood lately.
2. Only steam on the kitchen
gas stove when your wife is away; preferably for at least a day. This
avoids two problems; "What are
you doing with that damn thing in the house?" and "That's my
turkey roaster you turkey!"
Back to Front Page