Post 5 in a series describing the design and construction of a dining table, the construction of which is largely based on a Ming Chinese side table. Previous postings, described the nature of Chinese classical tables and detailed aspects of the joinery employed in these tables, at least in so far as the triple-mitered connections employed with corner leg tables. The previous post explored a unique Chinese side table which forms the main point of design inspiration for this piece I am about to build.
One point about the design parameters I am going to try my utmost to adhere to, in line with Chinese traditional practice:
- no glue unless absolutely necessary (and to be a reversible type if employed)
- no pegs unless absolutely necessary
- no lathe-turned pieces
So that sets the bar kind of high right off the bat. If I hadn’t seen so many magnificent Ming pieces to prove the validity and worthiness of that traditional construction, I might think such constraints were overly onerous, if not impossible to fulfill. The third item is no problem to avoid, but the first two are a real challenge in a piece like this.
Of course, the dining table brings with it certain other challenges not faced by the table we looked at in the preceding post. That table was narrow, and thus the top panel could be somewhat easily made from a single piece of wood. A dining table needs to be at least 38″ wide, typically, and making the top out of a single piece, while do-able, is not a simple matter. Given the unusual form of frame and panel construction I am seeking to emulate, given the fact that the frame and top in this piece meet along the table arris and not several inches in from the edge, and that the frame and panel nature of the piece is obscured, I have to employ material for the top that moves as little as possible laterally. That means employing, as far as is possible, radial grain material, aka ‘vertical’ or ‘edge’ grain wood. Most wood species move about 1/2 as much in the radial grain orientation than in the tangential (‘face’) grain orientation. However, pure V.G. material, free of sapwood, if I were to obtain it in a single piece for a table of this width, would have to come from a tree something like 8′ thick. Such trees are hard to find, and very few mills can in fact slice such wide material up.
Further narrowing down the criterion for this table a bit, I need to obtain a wood which is really dense and strong. The Ming originals were in such materials: huanghuali (a rosewood, dalbergia odorata), or zitan (Red Sanders Wood, pterocarpus santalinus). Neither tree is/was known for attaining a large size, and both trees were in short supply 200 years ago. Obtaining either of those woods is pretty much out of the question these days, at any price. I found a US supplier who could obtain Zitan, believe it or not, but the minimal order was 100 board feet, at $100/board foot. So, $10,000 just to get a bit of material. That dog, my friends, won’t be hunting.
The need for a hard and dense wood rules out virtually any of the softwoods. Softwoods are not a good choice for tables meant to last for a long time, especially if they are to be highly dependent upon their joinery.
In hardwoods, there are a few choices in larger sizes, however when one adds in a need for minimal sapwood and straight knot-free grain, we can rule out Walnut, and probably Elm. Most of the Elm in North America succumbed to Dutch Elm disease, so it is not so easy to find wide boards of that material. Honduran Mahogany can be found in wide boards, and is straight grained and knot free, however it is only in the middle zone of density and toughness.
A further constraint in this piece relates to the avoidance of using glue and thus I won’t be gluing thin pieces up to create larger blanks. So for the legs of this corner leg table I need to find stock in the 10/4 ~16/4 zone. Many of the hard and dense hardwoods, like Gonçalo Alves for example, cannot be obtained dry in sizes thicker than 8/4. So, the list of possible candiates slims down ever further.
As I mentioned, obtaining a piece of VG material in a width to serve as the entire top is pretty much out of the question. So the top will have to be made out of two or more pieces. The fewer pieces I can employ, the cleaner the look, while the more pieces used to form the top the wider the range of possible woods I can use due to the increased potential for obtaining VG material the narrower I go.
So, in a nutshell, to duplicate the fine Ming examples which inspire me, I am looking for a wood with the following attributes:
- dense and tough
- elastic rather than brittle
- available in decent widths
- available in thicknesses 10/4 and greater
- not highly figured
- mostly VG~rift grain orientation
- reasonably workable by hand tools
- capable of being polished to a high luster
- stable in service
And of course there are some price constraints on the material as well. I did find a stash of Indian Rosewood, cut in the 1930’s and kept since at a furniture factory in Austria, now at Hearne Hardwoods in Pennsylvania, with some pieces in the 18″~20″ width, and some pieces 16/4 thick. That would be a beautiful material choice, tough and highly polishable, but at $125/board foot it wasn’t realistic for this project.
In the end I concluded there is only one wood which is available to me here that fulfills most of the criteria, and that wood is bubinga (guibortia spp). I have worked with it before, most recently in the tsuitate project. Trees of the guibortia genus are part of the legume family of plants, and comprise some 13 species from Africa and, I was surprised to learn, 3 species in South America. Here’s what the tree looks like:
These trees occur in swampy or periodically inundated forests, as well as near rivers or at lakeshores. They are evergreen trees growing to 40+m. tall, with a trunk diameter of 1 m. or greater, often with a heavily buttressed trunk. Here’s a look at the foliage:
Bubinga, which may be either Guibortia demeusei, G.pellegriniana or G. tessmannii comes from Africa, the west coast more specifically, and to be most precise, from Gabon, Equatorial Guinea, Democratic Republic of the Congo, Nigeria and Cameroon. An interesting aside is that in Japanese domestic market they market bubinga wood as ‘African Zitan’ -obviously they feel it parallels zitan in certain qualities (though zitan is a dark purplish material). Though not a true rosewood, bubinga shares certain aesthetic qualities with rosewood, and is denser than many types of rosewood. So, that’s what I’m going with. It’s the best choice – it’s the only choice.
Even with the wood decision made, obtaining a suitable board was not altogether easy. Most bubinga that can be obtained in wider slab form, at least on the Eastern seaboard of the US, has been sawn for figure from figured logs. I wanted quiet, straight-grained stock, not figured. I wanted wide 12/4 stock. Bubinga is apparently quite easy to dry compared to many hardwoods and can be obtained, dry, up to 16/4. The dining table I am to make needs to be 8′ long, more or less, so that formed the minimum length under consideration.
After much searching, I did locate some down in Maryland. Then, when the time came to obtain the wood, the supplier suddenly was preoccupied with some other project, and has not returned my repeated phone calls,messages and entreaties to get in touch. This was beyond perplexing, since people don’t throw down $5000 or so on a stick of wood every day, especially in this poor economy, but for whatever reason this supplier has better things to do with his time than return my calls or sell me that piece of wood. So I have had to look elsewhere, and managed to locate another stick in Pennsylvania which is 4′ longer than I need but what can you do. I will make up the difference and buy myself 4′ of bubinga. I’m sure I’ll put it to use sometime. It is without doubt one of the most beautiful woods.
So that’s that. In the next post I’ll explore the design I have come up with for this table. Thanks for coming by today. Post 6 follows.