Springback when steam bending?


OK, so I'm contemplating my first excursion into steam bending, and I'm having a *lot* of trouble determining how much springback to allow for, when building my form. I've Googled, and read a couple of FWW articles and books, and the best I can find anywhere is general advice, e.g. the tighter the bend or the greater the degree of arc, the greater the springback -- well, duh! that would seem to be obvious. What I'm looking for, and am so far totally unable to find, is any source that attempts to quantify, however roughly, the extent of springback, so I have some idea of a starting point for my forms.
The specifics: approximately 75 degrees of arc, bending radius 43 inches, solid stock 5" wide by 1/2" thick, probably beech or maple, maybe poplar (it's going to be painted). I'm making a baseboard to fit into a curved alcove, so once it cools and dries, it's going to be nailed in place (which obviously will constrain springback to some extent).
I know I need to make the form somewhat tighter than a 43-inch radius, but does anybody have a feel for just how much tighter? 42 inches? 40? 36?
Any help would be greatly appreciated.
--
Regards,
Doug Miller (alphageek at milmac dot com)
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Not to try and talk you out of steaming your stock, but there are other options for baseboard. Nowadays, you can get manufactured flexible trim in many profiles. Some are made of resin, some are made from foam, and even mdf trim will flex quite a bit without a bending form. Even real wood trim can be bent and will retain the shape with some springback just by taking a ratchet strap and pulling the ends together slowly and tightening it a little more each day for a while. The latter is generally what I use in a pinch when trimming a house with tight curved walls. It often takes a good week or two before the board will bend to approximately the right curve, but it'll go. --dave

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Steam bending doesn't involve glue or laminations... and surely there's gotta be some data somewhere....
--
Regards,
Doug Miller (alphageek at milmac dot com)
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Doug Miller wrote:

oops! as soon as I read your response, I realized my brain must have been in neutral. DOH!
dave
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Doug Miller wrote:

Usually doesn't. I've made arched windows with steam bent laminations as the radii were very tight.
R
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I would think if you made it at 40" even if it sprung back more than you want it would be close enough. The wood will still have some bend to it . It has to be easier to bend an already bent piece of wood than a straight length . Just my thinking anyways. Jim

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I just steam bent a a piece of canary wood (arariba) 36"x3"x3/16". The application radius was 18 inches and I made the steam bending form with a 16" radius. The springback was considerable, and I had to use clamps (lots) when gluing the piece to the cabinet. If I did it again I would try a tighter radius for the form. Not sure how this relates to your application since almost everything is different!
If you can make the piece thinner than 1/2 definitely do so. The thinner the better.
I also made some curved doors by laminating 1/8" bending plywood. I think this would be a much much much easier approach for your application. There is zero springback, and you don't have to worry about the cracking and splitting that is common in steam bending. I've used unibond 800 and regular old yellow glue for the laminations. Both work well.
I had to order the bending plywood. I comes in 4'x8' sheets. It only bends in one direction, which you will have to specify when ordering.
Mitch
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Yes, I considered that, but decided against it, mostly because I'll also be using a table-mounted router to cut a profile on the upper edge of the piece. In bending ply, that would be almost entirely an end-grain cut, and I'm leery of the tearout.
--
Regards,
Doug Miller (alphageek at milmac dot com)
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

(it's
If it's going to be painted, why not laminate?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Probably more hassle IMO. I have to build essentially the same form whether I steam-bend or laminate, and if I laminate then I also have to do a lot of resawing -- since it's going to be 5" wide, that means the bandsaw, and *that* means multiple trips to the jointer and planer as well, plus the additional mess of the glue ... balanced against making a steam box.

--
Regards,
Doug Miller (alphageek at milmac dot com)
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Tue, 28 Feb 2006 01:58:54 GMT, snipped-for-privacy@milmac.com (Doug Miller) wrote:

Doug, there are a set of equations that will describe this but I couldn't put my hands on the key one - the one that I've found most useful relates to the relationship between number of laminations and percentage of springback. DAGS, you might find it.
Brief comments:
1) Successful bending requires compressing the wood cells on the inside of the bend - wood cells don't stretch very far and will break so the bending method must constrain the outside fibers while compressing the inside fibers.
2) the amount that you can bend (and the force it takes to bend) is a function of how much and how easily you can compress the inside cells. This is a function of the elasticity of the cell walls and the size of the cell.
3) The natural elasticity is best when the wood is green but can be temporarily increased with the heat introduced with steam but only if the material in the cell walls have not crystallized. There is a formal description of this condition which I don't remember but suffice it to say the main distinction is whether the wood has been air dried or kiln dried. Kiln dried wood is more apt to be 'crystallized and therefore more difficult to increase the elasticity.
4) The bending force/stress curves are very definitely NOT linear. There is a small part of the curve that is linear as the cells are compressed but then the force/stress increases rapidly as the cell approaches its fully compressed state.
5) Given 1) & 4) the amount you can bend a piece of wood depends on the amount that you can compress the inside cells versus their normal or slightly stretched state. Hence the amount you can bend is a function of the delta circumference between the outside of the bend and the inside of the bend. This is obviously a direct function of the delta radius (dR) which, of course, is the thickness of the wood.
6) Since the bending force curve is not linear, the effort to bend a 1/2 inch thick piece of wood is most definitely not simply 4 times the effort to bend an 1/8 inch thick piece of wood. Hence, laminations require FAR less force to bend the same thickness then a solid.
7) The greater the force, the more likely that your going to approach the steep part of the curve of the cells and end up breaking the bonds in the outer cells and the wood will fracture. Hence you are more likely to successfully bend laminations than solid wood.
Net: Since it is easier to bend a thinner piece of wood (smaller dR) and is a relatively easy thing to set up - especially in your case where the inside is hidden and you're not looking for structure strength. Cut multiple inside kerfs so that the effective thickness is much lower. Kerf spacing is a function of the radius - the dR over the depth of the kerf need to be less than the 1/8 inch thickness of the kerf. There was a pretty good article describing this technique and calculation in one of the mags last year sometime, DAGS it.
A good alternative to solid wood is to use MDF and kerf bend. For your 1/2 inch thickness, I would laminate an 1/8 thick piece of MDF to a 3/8 inch substrate (wood or MDF) that is kerf cut - this will smooth the outside surface and is a good hybrid between kerf/laminate bending.
Laminate bending is actually much easier than steam bending for a single piece, IME. You don't need to joint the inside surfaces - polyurethane glue will fill the small gaps. The equations, if you can find them, will show you that the springback is a function of the number of layer BUT the number of layers is actually a power (not a multiplier) in the equation so that the percentage of springback reduces dramatically with the number of layers.
Finally, if you do want to bend a solid piece, don't use kiln dried wood but use air dried or green (preferable).
Hope this helps, good luck.
TWS
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
One variable that many people forget is the time between removing the piece from the chamber and getting it on the form. A scant few seconds difference there can change the bend fidelity enormously. Since you said it goes *into* an alcove, I would think you want the radius to be a little large rather than too small. If the radius is too small then you have to hold the ends against the wall (forces parallel to the nails). If it's a little large, then holding it into the curve is managed by forces perpendicular to the nails at the ends of the piece. (It's similar to always making sure the seams in a large glue up are tight at the ends with a tiny gap in the middle.)
hex -30-
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Site Timeline

HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.