I'm building a steam box from BCX plywood, and putting it together with brads and glue. What would be an appropriate adhesive, that would withstand 212° Farenheit and a very wet environment? I know that the epoxies I use soften below that temperature, as does hide glue and I think TiteBond I. Franklin says their aliphatic resin glue bonds can be broken with heat and moisture, but I can't find any temperatures specified in their publications.
100% silicone cault (window/door stuff) is your answer. The stuff will take up to 500° which is why fireplace guys use it to "stick" door units and such to a masonry hearth.
Other thing is to use stainless screws. steel or brass fixings will not take the combination of heat and water.
In his book on chairmaking,Peter Galbert says he used polyurethane glue with cdx plywood and joined the long edges with a tongue and groove joint. I did this when I built mine and it works well. He also mentions putting a thermometer in the top of the box. It is a good idea. Then you know what is going on in the box.---Jack
Yes, leakage is necessary, but it is more efficient if the steam travels the whole length of the box. I have built a closure for the far end with adjustable vent holes to let the steam out. The steam will enter from the bottom, at the near end, which also has a door where the wood to be bent is inserted and extracted. Condensate will run out through drains in the bottom near the far end. The near end will be slightly elevated, encouraging the water to run out the drain and excess steam will be pushed out through the vents by the new steam.
I have Galbert's book, but I hadn't read that part of it yet. It is a very good book, and building Windsor chairs is yet another rabbit hole to fall down headlong.
So, there is an application at which Gorilla Glue excels. There are several reasons I really don't like using it, but its advantages in this case might just outweigh the disadvantages.
I would have used CDX, but the local Borg only had it in 3/4". Since the steam box needs to be 13' long, I decided to use the available 1/2" BCX instead, because it's lighter. The planks, etc. won't mind the marginally nicer box walls. It's pretty thin for T&G joinery, and using stainless screws to reinforce and clamp it would also be a problem, as I would have to predrill all the holes, or risk splitting the plywood. Brads, from a pneumatic gun, would be a lot quicker and easier, but I probably should look for Monel brads.
I'm planning to put four or five small holes in the top for thermometer access. With multiple probe points, I should be better able to tell how much to open the vents for optimum steam flow.
I don't know what Jack used, but I used Gorilla glue on my steambox. It is similar to Jack's, CDX plywood in rabbets. It has held up for over 30 chairs and 20 years. Make sure you put some dowels through the box strategically placed near the bottom to keep what you are bending off the bottom.
It's the only available glue that I know of that will not eventually fail at those temperatures and in moist, steamy conditions. It requires that the joints be well made and in good contact, as its gap-filling properties are not good.
Keep in mind, that if only a small portion will need to be bent, it makes more sense to make a box with holes or ports that allow the blank to stick in one side and out the other. This requires a lot less BYU's than heating a lot of wood that may not need bending, as well as no reverse bending where it is unintended, if there is a lot of end pressure under the straps.
I'm bending Atlantic White Cedar planks for the hull of a dinghy, specifically, this one. It seems like I've been working on its restoration for ever.
A Windsor chair is next on my list. I've built Chippendale chairs, but there are no similarities between them. Everything is totally different: the wood, the stock preparation, the design, the tool set, the joinery, the shaping, the assembly, the finishing, and the mind set. I'm still putting together my tool set: I still need a shaving horse, a tenon rounder, a kiln, and this steam box, and I haven't found out yet whether or not my bowl adze has the right geometry for hollowing seats.
I've done no serious steam bending so this is a naive question: would a box made from galvanized steel work? Any local HVAC shop could fabricate any length and any size rectangular or common round diameters very quickly and I suspect it would be a lot cheaper than plywood. I imagine ply ends.
The boat is a Classic Moth. They're 11 ft. long, and some of the planks run end to end, and this one has some significant curves. The only part of the hull that is not a compound curve is the transom. I think I was a bit generous with the length. I could probably make do with 11.5 ft. The box is designed to come apart in the middle, so I can use half of it for shorter projects. However, I think I have also overestimated how long it would need to be to handle a continuous arm Windsor, which is bent everywhere. I've also made replacement wishbone tillers for some of these old boats, which are a double ogee shape, and again bent everywhere.
You can make them out of almost anything. The downside of metal, is that it is a good thermal conductor, which can rob the chamber of heat unless it is well insulated. Of course even a plywood chamber can benefit from insulation.
It is all about getting enough heat to raise whatever volume of wood to around 210* all the way through. Plenty of things can prevent success. Just last week, I was answering questions for an old shop mate, now living high in the mountains of Colarado. I was prying all the details from him that a novice wouldn't consider, when it dawned on me that he lives at around 8000' elevation. At that elevation, water boils down around 190*,,, I've forgotten the exact number, but too low to soften the lignin.
I think the next time I build one, I'm going to use an insulating foam board. My old chambers are wrapped in loose layers of packaging foam for insulation that I had on hand.
Since I'm working with half inch sheathing, and I'm not even sure how it achieved the BC rating, I can't imagine getting tight fitting joints out of it. I guess I'm condemned to patching it back together with stainless steel screws and duct tape (or maybe chewing gum and bailing wire) when it starts to self-disassemble.
I suspected the same problems that Keith notes, but I don't have any experience either. The one I'm building is certainly inconveniently heavy, and sheet metal would help with that, but it has some other disadvantages. With wooden steam boxes, you just drill a pair of holes in the sides to run dowels through to keep your stock off the bottom. I've only seen them made from plywood and PVC pipe. PVC pipe has two disadvantages: it has a tendency to melt and sag, so needs extra support, and it is not readily available larger than 4" diameter. That's more than sufficient for Windsor chairs, but my planks wouldn't fit.
I've heard rumors that there is another type of conduit, made from a different kind of plastic, that softens at a higher temperature, but I don't know any more about it, especially where to find it.
Mario Rodrigues and Alan Turner at the Philadelphia Furniture Workshop also use 4" PVC pipe. OTOH, most of the other guys teaching Windsor chair construction build plywood boxes.