A Square Deal (2)

This is a series of posts describing the design and build of a square coffee table.

In the last post, I mentioned that the client had expressed a preference for Claro Walnut, and showed sketches of a preliminary design I had come up with which featured a 1.5″ thick solid table top intended to be produced from a single wide board of the walnut. The design was most of the way there, I thought, and the client indicated he liked the direction things were going. After a few further tweaks were made to the design, I realized that it was pointless spending much more time on it without being sure that I could obtain the appropriate piece of wood for the table top. Additionally, the legs on the table were around 2.5″ thick so I would need 12/4 (3″) thick stock for that. Ideally, the material for the table would all come out of the same log, if not the same board. Walnut tends to vary somewhat in tone, color and figure, for tree to tree, so my preference was to find a large slab of Claro walnut which would serve for the entire project. That was the ideal, however it was not being met by any of the boards I came across in my internet hunt. This is a medium sized piece of furniture at best, so beyond the piece of wood needed for the top, which was to finish out at 38″ square or so, I didn’t need a heck of a lot more material. Ideal would be a board wide enough for the top and thick enough for the legs. A mega-slab in other words.

Claro walnut, Juglans hindsi, is also called North California Walnut. It is commonly used as a rootstock for grafting on an English walnut stem. It endemic to northern California in a very limited area, with is only one confirmed native stand remaining. Thus it is listed as seriously endangered by the California Native Plant Society.

Fortunately, the tree has been planted elsewhere in California and Oregon as an ornamental and is noted as fast growing. So, supply of the wood from trees outside of the native habitat is not a problem, not that one would characterize Claro walnut lumber as abundant. A Claro Walnut of 100 years age can be over 4′ in thickness, and some reach up to 7′ and 8′ of thickness, particularly down by the flared root base. Locating wide slabs of Claro is therefore not terribly difficult, given the internet and the recent craze for making large waney edge slab tables from walnut, however, finding the right slab of wood for this project proved more of a challenge than I expected.

For starters, many of the walnut slabs have pronounced defects, including rotten sections, bark inclusions, remains of metal (as is often the case with urban trees), splits and cracks. Walnut tends to have a fairly thick band of white sapwood, and a lot of the walnut you see has been steamed as that process darkens the sapwood to look more like the blackish brown heartwood. Some people consider these defects as ‘character’, however I’m not in that camp. And the practice of dealing with a slab’s crack(s) by placing a row of butterfly keys in to control the movement has been done to death by now and I’ve never really liked the practice anyway. In truth, a butterfly key works like a single dovetail pin on both sides of the crack. The side grain of the dovetail head of the key bears against the end grain of the key’s mortise, and if there was appreciable spreading force in the crack the side grain of the key would be the loser in the exchange, getting compressed readily and sliding back. If the key is only installed shallowly into the substrate, as I have seen on numerous occasions, then the key does an even less effective job.

The above points are a bit moot in certain respects as walnut is an exceptionally stable wood so the cracks don’t tend to open much anyhow, and the keys might be best considered primarily as a decorative effect. And that’s fine – I guess the message that these decorative keys are transmitting is something along that of the tensions between wild and irrational nature (the crack) and rational controlling hand of man (the key). I speculate, of course. Maybe they just look ‘cool’ to some people. Nearly all people installing those keys put them in with glue, and some with screws as well, and I would suggest those factors are engendering a good portion of the connection’s mechanical resistance.

Sapwood, even when steamed to color match the heartwood, is something I avoid in my work. Sapwood contains sugars and is softer than the heartwood, so this is the portion of the wood most desirable to boring insects or fungi, whether for food or a home. Sometimes when the wood is dried the sapwood becomes less firmly attached to the heartwood and partially separates. Dried sapwood sometimes has a punky consistency. I came across a few walnut slabs where the sapwood had been knocked or chopped off – due to powderpost beetle damage in the sapwood. And while most of the vulnerability of wood to insect damage occurs after felling and during the air-drying phase, in looking at antique furniture in museums where one would expect that the wood used was originally air dried only, evidence of subsequent insect damage to sapwood portions is something I have come across several times. I cannot be absolutely confident that sapwood, even after kiln drying, will no longer be attractive to some insect or another years down the line, so I will continue to avoid using it.

Eschewing the sapwood means no waney edge either – in fact, with walnut I would expect to have to trim 5~7″ off of a board’s edge, given the generally less than straight grain, to obtain a piece free of sapwood. That means, to obtain a table top 38″ wide, I needed a walnut slab around 48″~50″ wide. Such slabs exist, but they tend to be fairly long too, like this one from Good Hope Hardwoods in Pennsylvania:

This board has some lovely ‘basket weave’ figure:

This board also costs $10,000. And even at that, given the goals I have for the one piece, a defect-free 38″ wide top, the above board shows cracks and bark inclusion in the section of the log from which a table top 38″ wide could be obtained.

There are makers out there who deal with cracks and voids in a board by filling the defect up with epoxy and grinding the epoxy down afterward. While this technique makes sense as an emergency repair, I don’t find it attractive in any way. It’s not a ‘feature’ for me. I would tend to want to patch a defect with a carefully-fitted piece of wood, though in general I would saw the board up to cut the defects out altogether. A large patch over a defect does not look that good either.

The idea, by the way, of constructing the table top out of 2 or more boards glued up edgewise was there in the consideration all along, however the ‘perfect world’ was to make that top out of a single ‘perfect’ board, so that’s what I was looking for first. If I couldn’t find that, then on to plan ‘b’, or ‘c’, etc..

As seen with the previous board, which might be called roughly typical in size for the bigger slabs of Claro Walnut one can come across, there is altogether too much material in that board for making just a coffee table, and working around the defects otherwise would likely mean a conversion of the board into the furniture piece of around 30%. It’s not the best.

I also looked hard for smaller boards that might do for just the top, and then obtaining stock for the legs and rails of the table from another board. One of the shorter slabs of Claro I came across looked like this:

While it might be possible to coax a table top out of that slab, especially if I made the top a bit smaller, say 36″ per side, the issue with the above board I don’t like is that is it primarily flat grain, and thus will tend to move more than a cut of wood with more vertical (radial) grain.

I also found this piece, possibly the only one which was going to provide enough wood for the top and was thick enough for the legs stock as well (but I wasn’t sure if it would provide the stretchers and rails, mind you):

The above slab was $4000.00. The grain is wavy, and largely flat-sawn orientation sawn from out towards the bark of the log, but with minimal cracking, no inclusions, and no sapwood. So, it has some pluses to be sure.

Since 9 and 10 foot thick hardwood trees are not a common feature on this planet, the possibility of a one-piece table top 38″ wide composed entirely of vertical grain was not on offer. I could however find slabs which had been cut a board or two up from the centerline of a larger log, and such a slab would have a large portion of vertical grain, which moves, typically, about half as much seasonally as does flatsawn material.

I kept looking and looking, and wasn’t finding the perfect board. The alternatives, as have been mentioned above, were:

• make the top smaller, say 36″
• make the top out of 2 or more boards
• make the top a frame and panel design which would allow a smaller board to work as the panel
• as mentioned in the preceding, and make the floating panel out of two or more pieces of wood

I considered the use of two pieces of wood, glued edgewise, to form the top. While many would jump immediately to the idea of using a ‘bookmatched’ set of boards, I don’t like to bookmatch as it makes for a construction where the grain of one board is traveling the opposite direction to the one to which it is glued. This can make planing the top along the glue line a hassle, and the light can reflect quite differently off of each board due to the grain direction, which partially mars the effect of the match. I prefer to slip match as this makes the grain in each board in the same direction. You can slip match a pair of bookmatched boards by flipping one of the pair over face-wise.

I found a few candidate pairs, like these:

 

Nice clear boards! Problem was, with the boards edged and connected up, I would not be achieving the desired 38″ width, and the boards were not thick enough to also yield the leg stock, so more boards would need to be acquired, increasing cost and increasing the likelihood that the color or figure might not be especially harmonious between the various pieces.

One can make a design and then go and find wood to make it happen – or not. Some designs are not going to be possible to execute in wood, and maybe a 38″ wide top was a design requirement which wasn’t achievable in the desired material in this case. One can also choose the wood first and then, possibly also as a result of inspirations/aspirations brought about by the nature of that piece of wood, construct something out of it, using what is available as a constraint. That might turn out to be what has to happen with this project, however I had proceeded first with design, then looked for the wood, not the other way around.

After a week of looking, I was beginning to conclude that the perfect board was not out there, and if it was, the slab I might find to purchase was likely to be on the order of 10~12′ long and cost in the $6000~10,000 range, depending upon how figured it was. In such a case it would mean buying far more wood than was necessary for just the coffee table. The client is also interested in a sideboard, however indicated a preference to proceed in a piece-by-piece manner, coffee table first. What I didn’t know, however, was: were it to come about that I would make the sideboard as well, would the client desire that the sideboard be made from the same wood/woods as the coffee table?

I sent the client a long email detailing the situation and options as far as Claro Walnut was concerned, mentioned the design ideal that affected the search, and the various steps back from that, and that a suitable board might exist somewhere out there (attaching a picture of the $4000 slab as the best candidate for the top found so far) but if a more ideal slab were located it may also mean buying a quantity of wood in excess of just the table. Such a quantity could serve for both prospective projects, though that would mean committing to doing both projects at the outset. I wondered what he thought of all that and what he might want to do, and needed this information before things could proceed much further.

While an ideal of mine might be to make the table top out of a single board, as I thought this gave the most aesthetically seamless and grand appearance, and precluded the need for using glue, I realized that such a desire might not be as important to the client. Maybe it’s just part of my special madness and no more? The client has to decide whether they wish to bring the resources to bear to realize certain outcomes, and what the artisan might consider important strictly on their terms might not fit into what the client desires and is willing to fund. I needed to know the client’s perspectives on these issues before design could proceed much further.

I mean, it was also a possibility that having a top out of Claro Walnut, one piece, in a 38″ width, was the client’s heartfelt and earnest desire, and if that couldn’t be met in some way then maybe the project wouldn’t happen at all. Who knows?

Now, I will mention also that, wanting to present more than an account of the narrowed range of choices I had discovered, options entailed by the design constraints I had set, I also thought it might be helpful to consider what the options might be if we looked at another wood species. The client had expressed a preference for Claro, and had a few pieces of furniture already which were made in that wood, but had also let me know early on that he was open to other possibilities as well. I had suggested Honduran Mahogany in an earlier communication, but that didn’t quite seem like what the client was looking for. And obtaining an 8/4 piece of Honduran Mahogany over 40″ wide for the table top was not going to be an easy task by any stretch. I did find another species which would fit the bill quite nicely – bubinga- and made mention of that in my email, along with a picture of a large bubinga slab I had come across. Yet another possibility would be sapele, and one supplier had some larger slabs of that, but they were just not quite wide enough.

I sent all this off to the client and waited to see what would come of it – hopefully not reading fatigue. In the next post I’ll share with you all the outcome.

Thanks for visiting the Carpentry Way. On to post 3