Hello and welcome to my blog! In this series I will detail the process of building a table and chair for use in an outdoor tea house. The seating arrangement will be partially exposed to the elements and so I’ve chosen to use Honduran mahogany, a wood that can survive the outdoors. The tea table will be similar to one I made previously for my sister’s family, of black cherry, while the chair is a new design with Shaker roots.
Being somewhat of a tinkerer, I can’t help but tweak designs as I build new iterations. Upon revisiting the tea table design, I’ve decided to make a few changes based on new experience along with advice having been given to me during the previous build.
The proportions of the table have changed to fit its new location and purpose, it needed to be scaled down in overall dimensions and yet I’ve made the table top a hair thicker. I’ve also decided to make changes to the stretcher to post connection, including changing the size of the stretcher to better mimic a timber frame connection. In addition to changing the size of the stretcher, I’ve moved from utilizing a wedged half dovetail to utilizing a wedged cog joint on the advice of Chris Hall.
On the existing table the top is battened using sliding dovetails, I will retain that form of battening. Currently the housings are cut into the table top and plugged at the exposed ends. While it remains a fun feature, I’ve decided that I would like to hide the housing under the tabletop and to do so I’ll rabbet a ledge, this additional step requires thicker material than previously used.
The winged supports are also attached using sliding dovetails, which were tapered originally. I will remove the taper and retain them in their positions with a locking mechanism; either a square peg or similar device to retain their positions centered along the edge.
Finally the joints at top and bottom of the legs are secured by a round tapered peg at the top and glue at the bottom. I will change that to secure both top and bottom with square pegs. If ever the table should need refinishing I want it to be able to be completely disassembled with ease.
To begin the process of building the chair, I’ve decided to prototype the design in basswood and white ash in order to work out specifics in full scale. I feel a chair is only good if it can provide a comfortable sit, so I thought it best to experiment before working in the mahogany that the final product will be produced in. The chair will need to be prototyped before the table is finished to ensure that the pair work nicely at their respective heights.
The work began by first locating Honduran mahogany, luckily I have a local supplier with very old stock that he’s been hanging onto for some years. I lucked out and found a board wide enough for the table top to be made without a lamination. I had hoped to minimize the laminations in this project, being that it is a scale in which that can be achieved. Certain aspects of the build will require laminations to work around ideal grain for those scenarios but a table top can be flat sawn without issue.
There is another three foot section to the board, already removed and set aside for the table top. I like to make some large cuts right away to let the material move if it will do so, and so I cut that even before loading it up to bring it to the studio.
One of the goals I’ve set forth in this project is to leave a plane finish on the chair seat, as with anything intended to be left outside I feel that giving that wood the advantage of a planed finish will assist in its long term survival. I will likely apply a film finish to the seat, but this will still further improve its ability to resist taking on water. That eliminates two important tools to chair making, scraping and sanding. Both of which are typically used to refine a seat profile. Instead I’ll be using tools specifically intended for a planed finish and leave a subtle texture in the scalloped, flat, and bowl cut areas of the seat.
The prototype started with 12/4 basswood, which I sawed into 2″ sections in preparation for creating a lamination. I used 12/4 material to begin with so that the chair seat would be rift sawn across its entire face. I’d like to minimize unexpected movements in the chair seat. The laminations were fairly critical, in that they need to be very tight fitting. As I cut the seat pan out any voids in the lamination would show up as large dark lines in the finished seat.
After the glue up, I jointed the seat surface flat.
The flattening work was done with the grain, utilizing a jack plane.
Then finishing up with a smoother. The chair seat was then laid out and cutout. On the basswood seat I’ve left a textured finish, applied with the spoon bottom plane. I enjoy the finish but I’m not sure just yet if I will also leave the same finish on the mahogany seat.
When cutting out the prototype I ignored documentation knowing that I would detail and document the process of cutting out the chair seat on the mahogany version.
The mahogany seat began similarly, starting with a lamination, joining up two rift sections to form the chair seat, then flattening it with a series of planes. After which I cut the seat blank to thickness with the same series of planes.
It was a fairly long process due to a scar in the material that needed to be worked out.
Once that was completed I was able to draw out my layout lines. Making some exceptions specific to this chair seat, which will have arms (the prototype will not have arms).
With the chair seat now laid out, it was time to begin the work.
The work begins with a tool called the chouna, or Japanese adze. A western adze is made to accommodate a mildly curved handle, where a Japanese adze is made to work with a extremely curved handle. The handle material is bent into a curve while still growing on the tree and left for some years to ensure that it retains its shape, after which it is cut and dried.
Normally one would lay out the posts and spindles, then drill them prior to cutting out the seat. I am not doing that here, the posts and spindles do not have a splay angle and so I can utilize my workbench surface to reference the rake angle of those holes. And I wanted to get right into the seat carving.
I purchased the adze from Yann Giguere of Mokuchi and spoke with Yann about the specifics of setting up a new adze, of which he provided me details on how to go about setting up. I setup the adze initially as a compression fit between handle and head, but in use I determined that I wanted a bit more ‘scoop’ in my cuts. Yann described to me that the reason for many to use a wedge was to manipulate the angle of the head.
So after using the adze for a while I decided that it was time to remove the head and cut a wedge. The wedge acts to tighten the head in addition to changing the angle of the head. You may notice that the wedge is applied in the direction opposite the force in which the adze is used in. It seems a delicate balance must be achieved to retain the wedge and use the tool. The adze is not used aggressively as one might think, not like one would take down a tree, but instead with precision like one might carve a chair seat.
In addition Yann provided some drawings on how to setup the adze. I was given an inside measurement of 210mm, but decided to start with a bit more than that to ensure I could cut my handle down to that number should I feel the need to change how it cuts.
The walnut sacrifice accounts for about 6mm.
Next I sharpened up the adze and was ready to go. The inside facet of the adze is flat, and so it’s fairly easy to sharpen.
The outside is curved, creating a ‘hamaguri’ or clam shell bevel. This is fairly well worked out from the maker, and so I only needed to continue working it with my Hakka and Uchigumori stones. Eventually the wrought will polish up a bit more evenly, but I’ll let that happen over time.
Finally a quick and easy cover was made for storage purposes.
The chair seat could then be cutout in a basic way with the adze. This step I did on the floor using my feet (wearing tabi) to position the work. It’s important to understand that I’m chopping down into the material and now back toward myself.
I worked as closely to my layout lines as possible. The adze, making its scooping cut, does not chop down aggressively into the material. This effect is perfect for a chair seat which will be cleaned up with a plane. There are no deep outside marks to work out.
Some of the obvious high spots will trimmed down with a fukamaru gouge (out-cannel gouge).
Leaving still a fairly rough surface.
A chair seat is basically six individual shapes converging into one another. The leg areas are two large radius half rounds and so they can be planed out to their finished dimension even before considering the other areas.
The plane used for this process is called the sori-ganna, or Japanese compass plane. I was able to get a wonderful sori on short notice from a fellow woodworker. This plane is an absolute wonder of a tool, beautiful kamaji pattern in the jigane, forge black on the ura, and a very hard hagane making for a plane that cuts for a long while.
I need to tune up the dai specifically for this task. being that when it arrived it wasn’t far from flat and obviously made for very long curves. The tuning began by making a kuchi-ire, or dovetailed key, for the mouth. This would close up the mouth and at the same time leave a very hard wear strip. The reason for doing this is that the wear strip right at the mouth opening was about to get very thin as I tuned the dai into the shape I wanted.
Next I worked a curve into the dai approximating the curve I had intended to cut, making certain that the plane would ultimately contact at three points; the mouth and the two extremes.
The plane was then used to hollow out the leg areas, creating two skate-board ramps.
The sori-ganna was then used cross grain to cutout the intersecting flat which would form the Y-shape in the center of the seat.
This flat will be further refined at a later point once another plane arrives, one which I will make a short dai for. The plane, while it is a ko-ganna, it is only so small. Rather than cutup the existing dai, I will instead chop a new dai using smaller material to make a mame hira-kanna.
The next step in the carving process utilizes a shi-ho-zori ganna, or Japanese spoon bottom plane. Joshua Villagas helped me out quite a bit with the terminology and when to use certain spellings of these words. ‘Ganna’ replaces ‘Kanna’ in situations where it is easier to speak through the combined words and zori replaces sori when it is after a describing term.
The dai on this plane must be where the majority of the preparation is spent, as it was very close to where it needed to be. I tuned the sold only very slightly to reveal the blade in a way where it protrudes the highest at the center then fades away to hide the corners as it go to the edges.
The blade, however, required substantial efforts on my part to maintain an even ura. Setting up the ura is critical to the life of the blade and so I put my effort in upfront when setting up a new blade.
The shi-ho-zori ganna is first used to rough out the bowl area, then I back out the blade and take 3/4 strokes going from one side all the way to the start of the opposing corner. A chip breaker exists for this plane, but its hardly needed here.
The bowl area forms the remainder of the seat, and if strokes are made full length and overlapping without starts showing in the middle of the cut, the appearance and texture is quite nice. Once the final plane arrives and I’m able to plane out the center, triangular shaped portion I will determine if the finish will remain this way or if a shi-ho-zori finish will applied over the entire surface.
Thank for following along, I hope that you have enjoyed this long introduction to my new project. I very much look forward to your comments.