Windsor Chair Resources

Windsor chairmaking tips-tools


 || Tip Index ||

Gluing up the upper half of the continuous-arm chair.

You will get more open time with the Elmers "white" glue as opposed to a "yellow" glue. PB 4/20/01

Make sure that all of your spindles, long and short, fit the mortises (sockets) in the arm with relatively equal pressure. If one is too tight, it will make assembly difficult...take your time when dry fitting the arm before glue up. Absolutely do not use yellow glue as it sets up way too fast, use Elmers white glue instead. SJM 4/23/01

Discussion of Glue Types

QUESTION: As you all know in chair making you don't always get all the joints real tight. And any round tenon joint is half short grain. Add to this the tremendous stress that chair joints are subject to and glue joint failures are inevitable I've been using yellow woodworkers glue with good results. But what about the long run? Is old fashioned hide glue any better? Some guys use nothing but epoxy. I heard of one guy who uses an automotive gasket sealer because its thick and never hardens??? Any comments on glue would be appreciated. JJ 4/25/04

REPLY: I've always used elmers white glue for all chair joint except glueing seat blanks together, where I use titebond yellow glue. The most important reason for me is the white glue doesn't get brittle when dry. It ends up just slightly flexible or rubbery, a good thing in a airdried wood assembly. It also seems to have a little longer working time and it's never failed. JT 4/26/04

REPLY:  There are a couple of ideas about glue in the chairmaking tips section that you might want to read. On a personal level I would recommend several glues as I have found them to be good in different ways. So..

The white glue has the advantage of setting up more slowly. It is convenient and ready to go out of the bottle. It is the glue I use to build up all the upper sections of my chairs.

Yellow glue("wood glue") has a faster set up time which can present problems if you are glueing a bunch of things at one time.

Both of these glues give the chairmaker a strong glue to work with and they are very inexpensive. According to the literature at 2 different glue suppliers, they claim the yellow glue gives a stronger joint than the white and water does not break it down as easily as the white. ?? I avoid the waterproof yellow glues for the same reasons I avoid epoxy: how do you repair a joint if you can not break down the glue?

Drawback(opinion): You can not repair a joint assembled with alphatic glue UNLESS you completely remove the old glue before reglueing. New alphatic does not grab old brittle glue. Repairs yellow and white glue joints on the under-carriage are very difficult.

Hide Glues: What the old timers used. Some builders today don't like the glue pot and the double boiling preparation. In my opinion, the hide glue from the bottle has far too much gelatin in it to keep it in the liquid form and this detracts from the overall strength when applied. In the hide glue family you will also find fish glue. This glue is sold as a surface preparation for windsor makers to "crackle the milk paint finish" when mixed a certain way with water. This is also an excellent glue to use in chair assembly. It dries slowly and the claimed tack is equal to the white glue. Hide glue is easy to repair. I know there is a tremendous amount of attention given to building chairs "under tension" and I use this method but joints can dry up over time and fail. Hide glue does not have to be removed to repair it. I have drilled a pin hole in the leg mortise and injected hot water followed by fresh hide glue and the joint has been good for another 10 years. No dis-assembly.

Epoxy: No place in this man's chairmaking shop. The stuff is great for boat building and other applications but it just makes repair work impossible. I would rather deal with a nail in the joint than epoxy. This brings tears.

You might want to ask yourself a couple of questions:

1. Do I make a chair that will never need to be repaired?

2. Do I want to make a chair that can be repaired in the future with glue... or do you want a nail hammered thru a leg mortise?

I can not say with absolute certainty, but, I do believe many of the existing windsors were repaired along the way if they were used heavily during their early life. Having made this claim, hide glue was the only glue available to builders till 1930-40 timeframe. Maybe this is a good glue to use?

Maybe using different glues in different places is the optimum solution??

I mark the bottom of the chairs that I have assembled with hide glue clearly so a person in the 22nd century will know what glue to use for the repair-- if they still have animal glues. Ha. Good Luck DO 4/26/04

REPLY: I still rely primarily upon what I believe to be the mechanical forces of wet-dry joinery, of legs stretched apart, and wedged tapered joints, so to me there are only a few joints where glue really matters - glued up seat blanks and arm paddles or knuckles come to mind. I use Titebond on the seat blanks, and I use a couple of #20 biscuts about 3/8" deep in the underside of the seat. I know they're a modern thing but they form a mechanical joint after they expand and I like that. On the arm paddles, I use Gorilla glue. I have found it to be the strongest glue for that particular purpose. Since I have it around, I use yellow Titebond for all the other joints. I don't seem to have a problem with it setting up too quickly, but I put chairs together a lot quicker than when I started, and I started with Elmer's white glue. BG-WKG 4/26/04

REPLY:  I appreciate DO’s remarks regarding his needs, but the question seems not to be what folks like to use and hope will work, but what glues are best scientifically for a windsor chair’s specific joints. Now I want to preface my remarks by saying I’m not a windsor chair maker and have never built a windsor chair of any kind. (I’m a period cabinetmaker and aerospace engineer, for what its worth).

The PVA glues, (yellow and white glues) all require high bondline or clamp pressure to generate full bond strength. The manufacturer of Titebond glues, Franklin, recommends 150-250 psi of clamp pressure minimum to develop bond strength, (depending on the specific glue and materials). I e-mailed Franklin and asked for test results for joints with little or no clamp pressure. They didn’t respond with numbers, but felt very little strength would result.

So any joints where clamp pressure cannot be applied (mortise and tenons, round or otherwise) are NOT good candidates for PVA glues. In the short term we may “get away” with PVA in these applications, but it won’t last, isn’t a strong bond, and is difficult to repair. For these joints a glue with good gap filling properties is required. There are only two such glues commonly available: Epoxy and hide glue. Additives can be added to either to get the open/work time and elasticity required for the Windsor chair’s unique assembly. I understand how this sort of alchemy has been deemed impractical by some builders.

Interestingly, the old timers belief that rough joints (like a rough sawn tenon) held better than two smooth surfaces is correct when the glue used is either epoxy or hide glue. The rough sawn, or irregular surfaces help to maintain the required bondline and stop the glue from extruding out of the joint. Good news for we sloppy workmen who produce gappy joints!

It is my understanding that some Windsor chair makers use tapered tenons in matching tapered holes in the seat-to-leg joints. (I use that joint too, but for other applications). While I have no way of knowing for sure, my guess is that a good rap with a mallet could easily generate the minimum pressure required for the use of PVA glue in this specific joint. Moreover, the line-to-line type fit makes this joint ill-suited to epoxy or hide glue (which will extrude out). You could score the tenon with a vee gouge to make a splined dowel type joint if you wanted to continue using epoxy or hide glue for these joints.

So just to sum up, anywhere you have a gappy joint or can’t produced a near crushing pressure with a clamp, you’re better off with epoxy or hide glue. If you are forcing a joint together, like a really tight set of dovetails or a tapered tenon, hide glue and epoxy will extrude out, so you may be better off using PVA.  AC 5/4/04

REPLY: This is fun. In your description of alphatic glue use in a "non-clamp" application one consideration you might be overlooking is the wedge. True the tapered leg tenon contributes to the pressure of the final fit in the matching mortise but.. the wedge driven in the tenon adds to the outward pressure on the mortise. I would guess equal to the reasonable force of a bar clamp on an edge joint.

Hide glue: Its still being used. You have to wonder: how has all of the 18th century joinery held up so long with this easy to repair glue? It seems like some good stuff after all these years. In most cases where I have introduced skeptics to hide glue they have been impressed with the ability to put a rub joint together with the hide glue with simple hand pressure. I have used this demo and come back the next day to break the joint at the glue line and it just doesn't happen.

Lingering Question: Each time I wedge a spindle and give it a good tap I wonder if there is any glue left in the joint? Well its not keeping me up at night I think the mechanical joint is "kicking in" at some point. For the chairmakes out there who have found a good thing that works for them, its hard to change. Why change when you have success? So we keep building and improving in our effort to build that chair just a little bit nicer each time. Its fun. So is the Chair Talk forum. DO 5/5/04


© Copyright Windsor Chair Resources, All Rights Reserved