The armbow of this design has a radical curve – two bends that are more than 90°. The original chair used curved branches to create the armbow. I haven’t been able to find a suitable curvy branch for this design, so I was faced with using flat planks with curvy grain.
I sawed out the arm parts from some oak that curved around a knot. But it wasn’t curvy enough to make me happy. There was too much short grain in the assembly to give me the confidence to use the arm.
So I ordered some cold-bend hardwood (sometimes called “comp wood”). I’ve worked this stuff for more than a decade and know what it’s capable of. My rationale here was to use what I had in order to avoid short grain – that’s is what the Welsh chairmakers did. They used curved branches to avoid short grain. I have comp wood.
My box arrived today and I knew immediately something was wrong. The wood felt dry and warm – usually it feels cool because of the moisture in it the wood. There was a small hole in the bag, which might have occurred during transit and dried out the stick.
I decided to give it a try anyway. The wood wouldn’t bend around the form while it was cold, so I put it in the steambox for more than an hour. That made it more pliable, but then it split open along the grain as I pulled it around the form.
I was about to go for a walk in the woods to look for some curved branches. But Brendan Gaffney talked me into using a bent lamination. Working together, we sawed up some dry oak into strips about 0.10” thick and then glued them up using hide glue.
Brendan also showed me how to use multiple strips of Masonite as a flexible clamping caul along the outside of the bend. I’ve never done that before (my fly is still open at this point, by the way), I’ve always used a metal strap covered in duct tape as my caul. The Masonite cauls worked quite well.
We needed 24 laminations for one arm, and it went together without a fracture.
So tomorrow I’ll do the second armbow and try not to expose myself to everyone during the process.
Katherine the Wax Princess has finished a new batch of soft wax that is now for sale in her etsy store. For those of you who missed her last batch, there have been a lot of changes to the way Katherine makes, packages and ships her soft wax. Read this entry for details.
The short story is that she switched to glass jars with metal lids (which are coated on the inside to inhibit rust). We switched to a waterless production process. And you get more than twice the volume of wax in a jar. Oh, and we have special packaging for the glass jars.
The new glass jars are sweet. The screw-top lid keeps moisture out and feels more secure than the friction-fit tin lid we used before. And the waterless production process ensures every bit of the soft wax is usable.
This might be Katherine’s last batch for 2018. She’s deep into applying for colleges, working on her paintings and pottery and continuing to become a budding movie critic.
FIG. 177. Bending an unsegmented felly. Kodavere, Assikvere village. Photograph by A. Univere, 1942. Photocopy 1146:4.
This is an excerpt from “Woodworking in Estonia” by Ants Viires; translated by Mart Aru.
Unsegmented felloes in Estonia and Latviawere always bent from ash wood. In Russia too, ash was occasionally used, although oak and elm were preferred. The Assikvere wheelright used ash also for the hub and the spokes.
FIG. 175. Bending blocks: In the foreground, bending block for wheel; above, pushing pole; in the background, bending blocks for two rear wheels; near the fence – bending block for sled runners. Kodavere, Rüütli village. Photograph by author, 1948. Photocopy 1105:92.
Here we will describe the process of unsegmented felloes as it was practiced until recently in Assikvere. Young ash trees had to be used for bending, the thinner ones yielding four boards and the thicker up to 12 boards. In order to make sure the suitability of a tree for felly making, a notch was cut into the trunk with an axe. If the piece of wood came out whole, the tree was no good for wheel making. When cutting wood for the front wheel, the length of the trunk had to be 8-1/2′ (2.5 m) and for the rear wheel 10-1/2′ (3.1 m). The boards were cut with an axe and wedge and then trimmed and smoothed. The width of the board before bending was about 2-1/4″ inches and it was 1-3/4″ – 2″ thick (5.7 cm x 4.4-5 cm).
FIG. 176. Top view of bending block: a. Pushing pole, b. Wedge, c. Holes for wedges; d. Holes for joints in bending block; e. Hole for rear wheel bending. Kodavere, Assikvere village.
The green boards thus chosen were steamed or heated for about two hours. Bending on the wheel bench was done while the boards were still hot. A large wooden circular mold was used for the front wheels and a somewhat smaller one for the rear wheels. This circular block of wood was attached to a cross-like base (Figs. 175, 176) to which the end of the board was secured beforestarting. Bending was carried out by three persons (Fig. 177); they were described humorously, according to their functions as the “pushing boy, wedging boy and pulling boy.” The pushing instrument, or the bow, is a pole between 6-1/2′-9- 1/2′ (2-3 m), where it is mounted on a metal pushing pin. The latter is held in position by a strong wedge. In working the bow (or the pole), the pushing boy forces the board round the bending block with the aid of the wedge. The operation has to be performed evenly, without moving thepole up or down. The wedging boy moves the wedge as the board is bent, and inserts it into the appropriate hole in the cross-like base of the bending block. Sometimes an additional wedge is driven in for firmness. The pulling boy goes infront and guides the board, making sure of the board’s direction round the block. When the circle is completed, the pulling boy and the wedging boy join the ends with two pieces of birch. An additional board is now used to join the ends of the bent felly from the inside, it having been previously steamed or heated together with the board for the felly proper. It is a job of the wedging boy to hold the ends tightly while the pulling boy attaches the inside joining board. Only then may the pushing boy release the pressure. During the bending the ash board may often “crack with the inner ring,” i.e. , along the inside of the annual circle. If the crack is thin and the wood remains sufficiently thick it is still fit for use.
FIG. 178. Bent unsegemented felly. Saarde, Saarde village, Photograph by A. Järvesoo, 1938. Photocopy 904:102.
Before proceeding with processing the curved unsegmented felly (Fig. 178), it has to dry in a room for at least one week. After it has dried the ends are sawn off, so that a space of about 1/2″ (1.2 cm) is left between them. They are later pulled together by the smith with an iron band.
FIG. 1. JACOBEAN GATE-LEG TABLE IN OAK WITH ELLIPTICAL FLAP TOP MOULDED AROUND THE EDGE The working of a shaped moulding is not difficult, but clearly the normal method for straight work, that of using a moulding plane, is not practicable.
The majority of woodworking operations are fairly obvious and, given practice, present no special difficulties. For example, to plane an edge straight, work a through rebate or groove, or use a moulding plane calls for little beyond care and practice. But what happens when the edge, the rebate, or the groove is stopped at one or both ends, or when the edge to be moulded is curved? Or how is a thin panel which has warped to be planed? These processes are seldom dealt with in text-books, but they are much more difficult to deal with, and in that sense are more important. We hope in this short series to show some of the ways in which these tricky jobs can be done. We begin with the case of a circular or elliptical table top around which a moulding has to be worked. There is nothing specially difficult about it, but clearly the ordinary method of using a moulding plane as for a straight edge is impossible
Whatever the shape of the top, it must first be cut out and trued up. In the case of a table having flaps the parts would be hinged first and the shape marked afterwards. They should then be separated whilst being sawn and trued up. Once again they should be put together to enable any unevenness of the joints to be corrected, and the line of the moulding marked out with the gauge.
Marking. The procedure varies to an extent with the moulding. In the example shown in Fig. 1 it takes the form of a bold torus, and a line marking the inner quirk of this must be put in with the cutting gauge. If the top is circular the line can be put in sweetly by fixing two slips of wood across the fence of the gauge as shown in Fig. 2. A touch of glue and two fine pins each will hold them. Some workers fix them permanently to the back of the fence and reverse the latter as required. Note how the slips bear against the edge. This cannot be done when the curve is not circular or when it is serpentine because the cutter would mark at a varying distance from the edge. Fig. 3 makes this clear, and shows how the cutter would reach in farther on the acute part of the curve.
The only way on an edge of this kind is to use the normal cutting gauge and keep it up to the edge as far as possible. It is inevitable that it will tend to run off in parts, but this will not matter because it will only cut into a part to be cut away later. The important point is to cut through the grain so that the latter does not tear up when the moulding is being worked. The crossgrain is specially important.
Use of Fillister Plane. The first part of the waste to be removed is the rebate shown black in Fig. 4. The ideal tool to use is the fillister plane. Set the fence so that it cuts just short of the gauge line, and the stop so that it ceases to cut when nearly down to the depth of the quirk. It is impossible to use the tool with the same accuracy as when working against a straight edge, but by giving the tool a turning movement as it is pushed forward it is possible to follow the line with reasonable accuracy (see Fig. 5). One word of warning when moulding the parts of a flap table. Do not carry the plane right through to the far edge straightway. It will only result in a splintered corner. Instead chisel off the end as in Fig. 6. There is then no difficulty about taking the plane right through except when nearly down, when special care is needed. It is often better to cut in from the other direction with the shoulder or bullnose plane.
If the fillister plane is not available you can use the ordinary rebate plane with a strip of wood nailed beneath the sole to act as a fence. This will have no depth stop, and the depth will have to be gauged round beforehand so that you know when to stop.
As much as possible of the waste is now removed with the plane, and, as a guide to making the whole even, two pencil lines are put it as in Fig. 7. The ordinary smoothing plane or block plane can be used to form the chamfer shown by the black portion in Fig. 7. Afterwards the corners can be taken off with either rebate or block plane, whichever is available or the more convenient.
The section now approximates to the required form, and the final shaping is done with the scratch-stock, a special cutter being made as shown in Fig. 8. Note that it leaves the lower edge of the moulding uncut as this has to provide a bearing surface for the notch of the scratch. A piece of old saw blade or an old scraper makes an excellent cutter. It is filed or ground across square, this enabling it to cut in either direction, an important point since the curve makes the grain awkward. The narrow part of the scratch-stock should be fairly long, so that by pressing down there is no tendency for the scratch to drop over at the edge. It is quite hard work using the scratch as considerable energy is needed, but a perfectly accurate moulding can be made.
If preferred the lower corner could be dealt with in the same way, but as the amount of wood to be removed is only small it is usual to work a chamfer all round up to pencil lines as in Fig. 9, and remove the corners.
To finish off two glasspaper rubbers are made. They are shown in Fig. 10. Note that the upper one has its side at an angle so that it fits against the quirk of the moulding. Start off with middle 2 glasspaper, and continue rubbing until all corners are removed. You will have to keep changing the position of the glasspaper so that an unused part comes into use. As the rubbers overlap an unbroken shape is produced. Note that the shape of the rubbers is a trifle flatter than that of the moulding. There is no need for the rubber to be curved along its length. When satisfactory finish off with No. 1 glasspaper.
“Readers of my column will know that I never make two chairs alike. Numbers of sticks, their spacings and length, the size of the seat and its shape, angles of stretchers, type of arm either steamed ash or solid wood, colours etc… the combinations are endless. I have gone to great length never to let anyone make a measured drawing, I just pluck the shapes out of the sky as it were. This is a reaction to a lifetime spent making things to others’ designs. I do however keep detailed measurements and photographs.
“This leads me on to gripe about some of the woodworkers I come across. I hope you will forgive my opinion. When I talk to readers or get letters it often seems to be about the petty cash of woodwork (technical points about dovetails or getting joints to fit), but rarely about shape, proportion or colour. I don’t think joints are that important. I would prefer to see woodworkers look at the total picture, is the piece they have just made beautiful, will it hold together, will it do the job it was made for?
“Woodworkers don’t buy my chairs, but they spend ages looking at the details of construction and then frown disapprovingly. They want engineering perfection. People who buy my chairs do so for two main reasons. Firstly and by far the most important point, they buy because they like the look of them. Secondly they buy them because they like sitting in them. They rarely inspect the joints. They think they look good, they think they will do the job they are made to do and even though the parts don’t fit particularly well, they are strong enough!”
— John Brown, Good Woodworking, June 1994, Issue 20
