A Ming-inspired Cabinet

Back in October of 2012 I started considering a sideboard design for our household. In that initial foray into that area was looking for design direction from Japanese kitchen storage cabinets, or Mizuya. Post 1 in that series can be found here.

I did many hours of drawing work on the Mizuya. By Post 10 in that series I had most of the framing details worked out and had even designed some copper drawer pull hardware for the piece.

I then took several months break from the design work and attended to other projects, and in that interim gave the project further thought. I came to see that the Mizuya pattern was not going to work so well. I had concerns about the weight and volume of the cabinet, the fact that it sat on the floor upon sills, and other matters. I decided to revamp the design, and move away from the Mizuya pattern for the most part. a ‘sideboard’ was hence the more useful way to refer to the piece. Then started a short series examining the possibilities in that regard, starting here.

By December of 2013 (see post here) I had come up with a design of two stacked cabinets, taking my cues from a number of different Chinese Ming pieces where a wardrobe sat upon a stand. There are no Ming kitchen storage cabinets to draw upon, it would seem, so the closest match I could find in terms of volume and storage function were the demountable wardrobes.

Post 3 in that series, published January 2014, saw me getting closer to a design I was happy with, however I then put the whole project on the back burner as I had other projects to do. The ‘Square Deal’ series of posts followed, where I built a pair of square tables for a client in Los Angeles. For that project the client chose to purchase an enormous slab of bubinga, 50″ wide, 3″ thick, and 16′ long, and half of that slab was consumed in the making of the two tables.

Then followed the kabukimon build, and then I took a couple of workshops in Japanese metalwork with Ford Hallam. Since then I’ve been back in my shop working on the Chinese wheelbarrows, number 3 and 4 in that series of pieces, for the artist Jeff Koons.

And, I have returned to the drawing board in terms of the sideboard. My west coast client has indicated he wishes to proceed with making a sideboard, and the remaining 8′ long slab of bubinga will serve as the basis for most of that construction. That I’ve been working on my own sideboard design for a couple of years has proved to be invaluable, however the client has his own particular storage requirements, and the piece I am designing reflects those imperatives. I do plan to make a sideboard for myself at the same time however, and it will be close in design to the one for the client, with adjustments for the particular storage requirements in our house.

In the 2 years in which I have explored the design for a sideboard, I have long struggled with two aspects. One is finding a way to visually diminish the mass of a tall cabinet. The other is finding a satisfactory way to break the cabinet into sections so it may be more easily moved and installed. I have found these issues rather tough nuts to crack.

For the volume issue, the situation you face is that once you have decided that the sideboard is going to be more than table-top height, @34″, the upward limit from there is going to be the ceiling. There’s lots of room to play with, but, how high to go? The higher you go, the more massive and ‘looming’ the cabinet appears, and accessing the tallest shelves gets to be an issue once you go much past 6′ in height. Stopping the height at 5′ seems barely worth it – you may as well keep it down at table top height.

One solution you see with Chinese Ming wardrobes is that the sides of the cabinet taper inward as they rise, however as soon as you do that you run into struggles with hinging the main doors, or, if there were to be sliding doors or drawers, their fitment to the cabinet sides becomes more complex and even awkward. As I tend not to shy away from complex construction, that matter was less of an issue than the door hinging issue. I just wasn’t seeing a good solution to that, so tapering the cabinet was out.

I also now had the guiding hand of the material itself – a large slab of bubinga. My initial thoughts were that since the board is rather special, as such a glorious width, it would be optimal if it could be employed in some way to celebrate that aspect. What comes with that however is the risk that re-sawing such a slab of wood for thinner wide panels may result in unfortunate outcomes, as you may end up with a bunch of unusable taco chips, if you get my meaning. Still, worth exploring.

Long time readers of this blog will remember a bubinga dining table I made using a fairly unique technical approach – this build was detailed in the Ming Inspiration series of posts. That build ran 50 posts, and in fact used a slab of bubinga from the same tree as the slab I currently have on hand. When I re-sawed from that other slab for table top panels, I cut 3 slabs at 1″ thickness to start, then worked two of them down in thickness over an extended period of time to 0.625″ thick. One of those slabs behaved nicely during that process, and one did not: eventually it warped beyond my control and had to be scrapped entirely. Fortunately, I had the third piece, which I was able to work down and keep within the bounds of flatness. The experience has left me rather wary of the re-sawing process with this wood however.

During the coffee table build in the Square Deal series, the 1.5″ thick bubinga slab had to be very carefully monitored and attended to for months, as it did evince a propensity to move at times. I had some definite points of worry along that journey, however I kept it flat in the end, also employing stress relief kerfs on the backside of the slab to help. Ultimately, locking the table slab into the support framing is what will keep it penned in, so to speak.

The nice thing about a slab top is that it displays the wood to greatest effect, with the least amount of distraction. A slab door however is not a practical option for a cabinet door, which you generally want to keep as light as possible. The risk of the slab warping is ever present, and unlike a table top, there is less of a surrounding connected framework to keep movement in check.

If you look at some of the best pieces of American made furniture from the 1750~1850 period, like the pieces from the shops of Townsend and Goddard, you will come across tables with a folding top used for playing cards. These tops were invariably made from solid mahogany, one of the most stable woods there is. These tops also invariably warped, so that when folded the table leaves do not lay flat upon each other. I would say that a frame and panel top would perform better in terms of keeping flat over time.

To make a door light, frame an panel is the way to go, however as mentioned above, this is not the best way to advantageously display a wide piece of beautiful material. So…how to have the beauty of the slab look with the stability of the frame and panel construction system? Some may immediately conclude that veneered construction using a plywood or MDF substrate is the obvious answer, however it isn’t an answer I will ever choose.

The answer for me lies in that framing system I used on the Ming-inspired table a few years back. It does make for a more complex bit of framing, but I know from experience that it can be done, so, gol’durn it, that’s the way I am going to go.

——

The first iteration of the ‘2015 design’ looked like this:

I was imagining using large Brusso knife hinges for the main doors and other fairly discrete hardware. The cabinet would be in two parts, a stand and a large box.

The first revamp came from considering the stand. Just like a sawhorse, a much stronger and stabler design it to splay the legs of the stand. This also gains a more ‘planted’ appearance. The compound joinery aspects of the splayed lags with through-tenons is something I have tackled many times in the past, so it was no hindrance at all. In fact, I relish doing that form of construction as its virtues are so clear to me.

An aspect of earlier designs which I was interested to retain was the shippō-gumi lattice pattern of overlapping rings. I wanted to explore if that motif could be incorporated into the cabinet, and this was the result:

The lower stand, I might add, echoes aspects of the coffee table designed and built for the client last year, with pillow blocks and stretchers with jogged-up central sections. The latticework section though does ruin the aesthetic of uninterrupted wood grain running down the doors, and makes for slightly weaker door frames as well. I dropped this idea, at least for the main doors, as a result.

The massive volume of the cabinet was still problematic to me. I tried rearranging the doors so as to put the smaller doors up top:

That really didn’t get me much further along. I was still stuck on this design apect. Sometimes you just have to wander around in circles mumbling incoherently until you stumble upon a solution or inspiration strikes. Or hit the bottle, resort to prayers, etc..

Tapering being problematic, I tried the idea of stepping in the top portion to reduce volume:

At this juncture I had also introduced a central divider between the main doors. But I still wasn’t content with the design. It still loomed too much even with the diminished top and it looked far too ‘box-like’. I needed to finish off the top in a different manner, however I wasn’t looking to crown molding or a ‘bonnet’ pediment with turned finials, eagle in the middle with outstretched wings or anything like that. The client also expressed a desire that the cabinet not have large portions that were simply decorative.

More consideration led to a decision to axe off about a foot of height and finish the top with what the Japanese would call a ‘Chinese’ gable, however in the west the architectural term would be ‘cusped’ gable:

At this stage I also simplified the framing of the lower stand, removing a ring of stretchers. The cabinet is now about 6′ tall instead of 7′ tall.

In the design above I have also gone to a traditional -well, ‘archaic’ would be a better term – method of hinging doors, in which the hinge side stiles are extended top and bottom with axle shaft extensions which are captured in the framing top and bottom. While the elemental simplicity of this type of door hinging, seen originally on Chinese temple doors and also on Ming wardrobes, has long appealed to me, its functionality was admittedly less than optimal. Hinging doors in this way generally means that the door can only open to the point at which it runs into the surrounding framing – sometimes this is less than a 180˚ door swing in fact.

Considering the design afresh, however, I realized that if I moved the axis of rotation outboard, ala Brusso ‘L’ shaped knife hinge, the door could be made to pivot a full 270˚ out. My solution therefore involves attaching an outer cylindrical rod to the door’s outer arris, like this:

This solved the problem except for one annoyance: the width and depth relationship of the cabinet made for doors which were so wide that when swung back and around they ran into the wall before they could lay flat against the cabinet sides. That was not the end of the world, but I desired a cleaner solution.

This problem was surmounted when I reconfigured the main doors so as to be bifold doors. I also removed the central stile so as to give uninterrupted access to the cabinet interior:

I’ve also reinforced the outer corners of the hinge framing with a bit of iron (or maybe bronze) strapping. This is a nod to Japanese tansu, though in those pieces the strapping often has more of a decorative function. There are some  Chinese and Korean pieces with metal corner reinforcement, so the examples are present in the old pieces. I wasn’t looking to add decoration, just strength.

The bifold doors will be secured by adapting bronze cremone bolt locks. As the doors are hollow, I plan to mount the hardware inside the doors, with just a handle or keyed lock on the outside.

And what of the inside of this cabinet? Well, we’ll get into that in the next post.