Dark Chocolate and Sponge Cake (2)

I was planning to start talking about the design of the cabinet, however there’s another issue occupying my attention at present. Designing and building is an interesting process, and you can never predict everything, despite making all efforts to do so, when it comes to solid wood. This is especially true when you are buying wood at a distance and relying upon photos and descriptions of others.

When I designed the cabinet which forms the focus of this build, I drew it with a frame and panel system, and did so employing quite wide panels. Wide panels, after all, offer a clean and uncluttered look, and typically make for stronger construction as compared to using divided frames with multiple panels.

Initially, my thought was to use curly shedua for the panels and mahogany for the frame, however it proves impossible to obtain shedua wide enough, and I avoid glue ups generally, so that plan hit a snag. Shedua may yet feature in this build however.

Then I found some exceptionally wide 5/4 (1.25″ / 31.75mm) slabs of Honduran Mahogany, at a nearly unheard-of 48″ (1219mm) width. These appeared to provide all my panel requirements, and were thick enough that the prospect of re-sawing the material became within the realm of consideration. Since the alternative to re-sawing is to plane the material down, thus putting more than half of the stock up the chip collection system, I was keen to see if I could obtain as much as possible of what I wanted by re-sawing.

As mentioned in the previous post, the risk to re-sawing is that material un-evenness and movement after cutting may mean that you cut a precious board in half along it’s thickness only to end up with nothing usable.

I realized that, for certain panels on this cabinet, namely the top panel, a board with a small centered band of flatsawn ‘cathedral’ in the middle would be the most ideal aesthetically, given the options. The required top panel was too wide for it to come from a quartersawn cut off of a 48″ slab . That cut would, to have enough width, include a portion of flatsawn at one edge, and that pattern of figure would look visually unbalanced. The alternative to that approach was to cut off a chunk of 48″ wide material to the required length and then slice off  10″ from each edge, thereby leaving the middle portion with centered middle cathedral of flat grain. That could then be planed down to produce the panel.

Since every slab of wood has certain potentials, which diminish with every cut made, I figured it made more sense to obtain a narrower board of equivalent board footage to the slab remnant and use it instead of the 48″ wide slab remnant.

I looked first to a local hardwood yard that has been holding onto a pile of about 2000 board feet of Honduras Mahogany, all 4/4 (25.4mm) thick, and in various wider widths. They’ve had this stuff for at least 20 years, and it hasn’t sold to any significant extent because they have kept the price on the high side, unchanged at $25/bd.ft for the last 10 years now. They know their business best I guess, but the last half dozen times I have popped by that yard in the past year or three I have been the only (prospective) customer, so I’m not so sure their business strategy is working all that well. I would have thought the idea of a wood yard was to bring wood in and then sell it, not warehouse it. Wood that sits in stock for extended periods is just going to cost them money, for the most part.

Anyway, they happened to have 4 boards in their pile, boards which were 29~30″ wide, and being 4/4, the only recourse with that stock would be to plane it down to obtain panels, which was easier to consider since it involved significantly less waste than doing so with 5/4 slabs.

They got the forklift out and we broke the pile of boards apart, mouse turds and mouse nest building material flying everywhere (the wood pile has been sitting undisturbed for quite a while it would appear), until I could get a look at the 4 boards. Unfortunately, not one of them was suitable. They all had too much flatsawn grain, a result of being through-and-through cut pieces a few slices too far up from the center of the tree. I was looking for a board which was one of those slices just above the heart center, which would keep flat grain to a minimum, and quartersawn grain to a maximum. I had hoped that yard would be the convenient answer to my problem, but it was not.

Then I looked over Irion Lumber’s website again, and found some suitable material, with the aid of a bit of phone consultation with the salesperson there.

I decided to buy two more slabs from Irion, one being the last of the 48″ slabs they had on hand, and the other a 36″ wide slab which, from the photos, has a minimum of flatsawn and a maximum of quartersawn grain. My thought with the 48″ wide, 10′ long piece was purely, “ain’t gonna see that again so buy it now and hold on to it”, while with the 36″ wide piece I saw it as a perfect swap-in for the 48″ x 88″ remnant I had from one of the client’s slabs, in fact it was a near-perfect swap on a board foot basis. The 36″ wide piece was 14′ (4.2m) long, and in order to save on trucking costs, I had them lop off a 4′ length from it, to keep the remnant board the same 10′ length as the 48″ wide one.

This wood made it, via Fedex Freight, from PA to my shop in under 24 hours, somewhat astonishingly. When I got a closer look at the 36″ wide board I found it was not quite what I expected. I had been thinking that, as it was a slice of a 36~38″ diameter tree trunk, it would have the center of the flatsawn grain portion more or less in the middle of the width. Instead, it was at the 3/4 mark, 27″ in from one edge, 9″ from the other. Possibly this board came from the bottom of the trunk where the buttressing is located.

So…my idea to obtain a piece for the cabinet top which had the centered cathedral was not to be, at least in terms of getting it from the newly-acquired 36″ wide material.

But…the configuration of grain did offer an exceptionally wide band of quartersawn material. I thought I could look to make use of that feature.

I felt that for one of the interior panels, the cabinet floor or the middle shelf, I could be fine with using 2-piece panels. These could be edge joined in various ways, with glue or without. They don’t have a significant structural function, they just divide or enclose space and are a surface for beddings to sit upon.

I took the 4′ long, 36″ wide piece and cut out an 18″ wide chunk of the quartersawn portion. I jointed one face and edge, then ran the stock though the planer to clean off 95% of the opposing face. From there, I could examine the run of the grain more carefully. After lining out the board to obtain the best grain alignment, and trimming with a saw, then re-jointing an edge, I obtained a 15.5″ (393mm) chunk, this being close to the tallest thing I could stuff through my Hitachi re-saw. It was a hair under 1.25″ thick, and I gauged off a pair of lines 9/16″ (14.2mm) in from each face along the edge, leaving a space for the saw kerf.

The boards did not move too much in the cut, and I was happy with the results at this stage:

The pieces are weighted as a precaution, not because they are warped or bowed.

As it turned out, the saw-cut I took was a hair fatter on one side than the other, and after dressing off the surfaces of both pieces in the planer, I obtained just two 3/8″ (9.5mm) panels – on one piece, I was lucky to do so, with the final pass to dimension taking off the last bit of rough sawn face. These will be fine at this thickness for the middle shelf of the cabinet, sitting in a frame about twice as thick as the panel.

However, I realized that the likelihood of obtaining two 1/2″ panels from this process was looking decidedly unlikely, especially when factoring wider boards, where I am wanting to obtain single-piece panels (like for the front doors, rear panels, and the top of the cabinet). And with wider material, I was faced with a greater likelihood that flattening out from any irregularities in the wood surface to begin with, and the potential effects of wood movement after the cut, were decreasing the chances I could get the material out from re-sawing that I wanted.

And even more to the point here was that wider boards meant no use of my Hitachi re-saw, which has 5 horsepower and stellite teeth giving just a slim 1.6mm kerf  – the other saw options, either powered or manual, would likely associate to a wider kerf and therefore less chance of obtaining the results I was after.

Speaking of re-saw options, after my last post a reader wrote me from Denmark and, in a long detailed message, described his experiences with both French frame saws and old Japanese mae-biki oga type saws, the ‘Whaleback’ form of single-man rip saw, and advocated strongly for me to use the maebiki oga. He was so confident that it was the best choice, and that the French saw was a lousy choice for the application, that he offered to send me one of his saws if I paid the shipping, on a ‘give it a try and see how you like it basis’. Incredibly kind! It’s on it’s way now, and I should have it in hand in the next day or two. I cancelled my order for the French frame saw blade and got my money back no problem.

However, to continue, after my experience with re-sawing that first plank, I was thinking I needed to look at plan ‘b’. However, the prospect of planing the stock down still caused me to hesitate. Was there another way?

Maslow talked about a Hierarchy of Needs. perhaps rather less formalized, I have my own, as applies to my work in wood, and here are a few worth mention, in no particular order:

  • that I use wood wisely, both in how I select it, process and in how the design is to maximize lifespan
  • that I design so as to limit wood movement via care in selecting material for purchase, and to cut, and in orienting grain in the piece
  • that I employ joinery primarily, with reluctant recourse to fasteners and/or glue
  • that I design to maximize the strengths of solid wood
  • that I use the tools at hand preferably over taking pieces to other shops to have tasks done
In the last post I joked about how I wished to employ solid single piece panels, and wanted to avoid having to do any glue ups with panels. I also wanted to avoid planing the wood down from full thickness if I could help it, while recognizing that it was the surest path to obtaining the desired stock.

Driving home from the shop, I reconsidered the situation in light of the re-sawing reality. and in light of the collection of material I had to work with, which was generous by most standards. An idea came to me, which I later sketched out on a scrap of paper, finding a way to use the boards as ideally as I could:


In a hierarchy of needs for this project, something had to give. I couldn’t obtain exactly what I had wanted in the design, like the one-piece door panels on the front so as to present a view of the entire tree trunk via the two front panels, without getting into planing material down. I didn’t want to revise the design into doors which have two panels each, and I didn’t want to do a glue up to make panels if I could help it, not on the most exposed panels in the cabinet, namely the front doors and top panel.
The things about these conflicting needs is that one can reflect and figure out which needs are more important than others. Given a choice between planing a piece of slab down and wasting 70% of it in chips, or making a panel out of two pieces edge joined, I generally would have to prefer the latter option unless I am sure that re-sawing will not work. Making wise use of material generally trumps the situation if the wood allows some design change options. However, if it were the case that minority of panels out of the set were to be planed down, rather than all of them, that was an acceptable trade-off if the reward was there.

So, I have forged a new battle plan, as it were. This plan results in the following:

  1. I take advantage of the fact that the 10′ long 36″ wide board has such a severely offset flat grain portion, and chop out a pair of 46″ long pieces to obtain the two door panels, purely from the wide portions of quartersawn available on that stick. To me, the benefits of the quartersawn wood outweigh the aesthetic of having the front doors display a tree trunk across the two panels. These two door panels will be planed down from full thickness.
  2. The 48″ wide slabs I have which are still full length, at 10′ (personal stock) and 11′ (the client’s piece) respectively, were not sawn by the most sophisticated equipment originally and if one tried to obtain a single piece, like a conference table, from the full length, then one would remove a huge amount of stock and be left with something like a 5/8″ thick top or worse, and that is if there were no movement after stock removal, which is unlikely. A wide top that thin would perhaps be a bit unstable too. These slabs therefore are not ideally suited to becoming giant tables or desks, and in order to make the best use of them I think they need to be cross-cut, initially into a 1/3~2/3 format. This will provide, with the 1/3 sections, material for the remaining panels, and leave, with the 2/3 sections, a comparatively less bowed/kinked plank, which, if it were to become a large table or desk, would be a thicker piece after dressing than if the piece were not crosscut.
  3. I will accept the use of 2-piece construction for the remaining panels, by cutting slab remnants so as to obtain pieces that are entirely quartersawn. If you consider the choice between taking a 48″ wide slab and cutting the 10″ flanks of quartersawn off, and then planing down the middle portion, which is about 50% flatsawn, to obtain a one-piece cabinet top panel, versus cutting 14″ bands of rift~quartersawn off both sides, and then re-sawing those into a pair of panels each, and accepting the 2-piece construction, I have to go with the latter option. It will move less and not be prone to developing splits over time as flatsawn panels can.
  4. In step (3) I am resawing slabs to obtain pairs of panels, and one set of these is to be 1/2″ thick, and I can’t get two of those from one blank. However, I could re-saw each of the two blanks slightly off the centerline, so as to be able to produce one 1/2″ board, and one 3/8″ board, from each slab. Then I simply pair the 1/2″ pieces together and the 3/8″ pieces together afterwards. Maximizes the use of these two pieces of quartersawn slab.
  5. The two 46″ long half-slab sections already trimmed from one of the slabs (see the previous post) were intended to become the front and rear door panels via re-sawing. One of those pieces however has a few bug holes and these holes go right through the plank, visible on front and back faces. I cannot use these for the front panels (and have a better option anyway in (1) above), however they could be employed for the back panels, and there is an option to either plane them down, manually re-saw, or trim to maximize the quartersawn portions, and then re-saw those to make a pair of 2-piece back panels. I’ll decide once I get into them further.
The above approach to cutting the various panels will hopefully leave the greatest amount of full width slab stock untouched, give me the nice 1-piece doors I wanted, and then 2-piece panels elsewhere, all quartersawn. I could choose a variety of ways of connecting the two piece panels together, including glue if desired.
Onto the battlefield to see what eventuates – hopefully the distance between plan and realization will not be too great. Here is the weapon of choice for the slab sectioning, all 190mm of it:


It’s a great saw that makes a very thin kerf.

Here’s the 36″ slab undergoing the first cross-cut:


After cross-cutting with a 96T blade, I swapped in a 24T blade for the rips:

After snapping a line, I let ‘er rip:

That was followed by more cross-cutting to produce the two front door panel blanks:

I then dressed the blanks down on the planer to see what I had with clean faces to look at, and this is the result:

As you can see, the right hand board has a curious bit of grain, which looks slightly like a bark inclusion but it isn’t, just a sort of darker oxidized piece of folded/distorted grain:

I think I am fine with the panels not looking perfectly identical or having perfect runs of grain- this is a natural material after all, and I think that the grain ‘curiosity’ adds something nice.

Then I took the two previously-sectioned slabs, formerly intended for the doors but now slated for becoming back panels, and gave them a clean up on the planer. Here’s the result after hitting target dimension for thickness (prior to re-sawing, I mean):

Unfortunately, the right side panel also has a hollow on the opposite face as well. Getting it clean means at least another 4 mm would have to be taken off.

I mentioned in the ‘plan of attack’ that I was uncertain as to what I might do to obtain the rear panels, however the condition of the stock dictates direction clearly here, as there really isn’t quite enough meat there to re-saw out the quartersawn bands and create 2-piece back panels. So, I will plane them down, obtaining single piece panels.

Then onto the task of obtaining the panels for the top and bottom of the cabinet:

Face jointing next of course:

Now that I could see the grain more clearly, I corrected the boards to obtain better alignment to the run of the grain:

Edge jointing followed:

That was followed by a rip on the tablesaw to obtain parallel edges.

Then the same rounds of re-sawing and planing were followed, and at the end I had a bigger stack of panels than at the beginning of the day:

All is not quite well however. One of the quartersawn blanks (for the top and bottom panel sets) had several insect bore holes, and these are fairly apparent in both sliced pieces after the boards were dressed to size. Bummer.

That leaves me needing another 1/2″ panel and another 3/8″ panel, both quartersawn. I have a 48″ x 88″ slab left. I was considering going and buying a piece of 5/4 or 6/4 (preferably) quartersawn mahogany, which, in a 14″ width and 4′ length shouldn’t be too hard to source. Then I looked more closely at the slab remnant, and noticed a sprinkling of insect holes there as well. So, the prospect of preserving that slab as a future potential table or desk top was less a proposition than I had been thinking it was. A single bug hole or two can be patched, but 8 or 10 holes, well, I think that is less worthwhile. So, I will sleep on it, but I think I’ll be slicing a piece of quartersawn material off of that slab remnant, and hopefully that will be all I need. It means that slab remnant’s potential future as a large single surface is not going to happen. I don’t generally build that way anyhow, so it is not a terrible outcome, merely a bit short of ideal.

It’s amazing to me that cutting the panels for this devoured so much material out of the two large slabs. I’m just going to have a portion of one slab left over. I was glad to be able to re-saw for a majority of the panels, and, if it were not for the bug holes, the plan had been spot-on and would have led me to the outcome I envisioned. but things sometimes do not quite go according to plan. At least, at the end of the day, there is enough wood on hand to realize the outcome, even if I cannot preserve as much slab as I had hoped.

All for this round – next post I’ll show some pictures of what I am making, as I imagine some folks might be getting curious. Thanks for visiting. Post 3 follows.

2 Replies to “Dark Chocolate and Sponge Cake (2)”

  1. I really appreciate the details that you go into about the challenges of stock preparation. It's a subject that often gets scant attention, especially with regards to the many compromises that wood presents, as you say: it is a natural material after all. This is much of a craftsman's / preindustrial approach, letting the material have a say in what the piece will be and the process to achieve its completion.

  2. Poto,

    glad you enjoyed the read. I enjoy the challenge that solid wood presents, even if it frustrates at times and requires changes of course when you thought you had it figured. 'Corralling sheep' is an apt metaphor.

Anything to add?

error: Content is protected !!
%d bloggers like this: