BCM Tansu Repair (III)

Here’s something new on my blog, as we near the 1000-post mark, namely hammer meets nail:


Hah! The above photo shows how the POC panels came out with the stain applied.

The trick I use to prevent any splitting of the board when driving in a nail so close to the end is to clip the head off of one on the nails and use it like a drill bit to pre-bore the nail holes in the board:

Some work remained on the second sliding door from this cabinet. It had a badly rounded outer arris on the tongue of the bottom rail, and that made the door not want to track cleanly. So, I rebated a chunk out  of the arris and prepared a piece of wenge as an infill:

The infill piece was glued and clamped, then later cleaned up by plane and a bit of stain applied:

The lower mizuya cabinet is more or less done, save for a reassembly of one of the doors. I need to bring some rice to the shop to use for glue, so that can wait until next time.

I set the lower mizuya-dansu aside and put the upper half of the heya-dansu on the sawhorses:

This is a wider and deeper cabinet than the other one, and unlike the mizuya-dansu, which is viewed on 3 faces, this piece sits within a closet and only the front face is exposed to view. Thus, where the maker took time to make the sides of the mizuya-dansu look presentable, with the heya-dansu only the front face received attention. The rest? Well, a little on the crude side, by any standard.

Cabinets like this remind me of buildings in a wild west town:

What you see on the front isn’t what you find in the back. It’s a stage set.

On the underside of the cabinet there is a stiffener added, presumably to support the floor boards mid-span. Curiously, the tenon shoulders are cut a wee bit back:

At first I was grinding my teeth thinking that they cared only about the appearance of a through tenon on the side and not making a sound joint – then I realized that you can’t see the cabinet sides, so the reason for the joint to be configured in this manner is beyond me. The other end is similarly gapped. Every time I look at it I shake my head.

The bottom of the cabinet reveals the dabo, the floating keys used to align upper and lower cabinet to one another:

With the cabinet on its back, it was obvious that some of the nails holding the ceiling boards on had given up the ghost:

The shelf support pieces are affixed to the cabinet side walls with nails, which were simply driven right through and clenched:

The proportioning of the frame stiles to the rails does not seem to have been well considered in terms of depth – note the offset between stile and sill:

Same sort of thing where the stile meets the upper rail:

In a similar vein, the main frame rails at the rear of the cabinet are smaller sections than those on the front. Again, since the side of the cabinet is not normally viewed, the disparities in frame part sizes is concealed.

I noticed also that the inside of the right stile was dadoed to receive a tongue on outer stile of the sliding door:

That’s a bit unusual, it seems to me.

The left stile, since it receives the rear-mounted sliding door, has but a little rebate to accept the door’s outer stile tongue:

Anyway, that’s a brief tour of this cabinet. Like the previous one, the main task was to renew the sliding tracks. With this cabinet, the tracks had remnants of track infill pieces, what are referred to as ume-gashi. They were trashed, but they were largely intact. I was able to extract these relatively easily with a chisel:

Ume-gashi are typically made from cherry, keyaki, karin, and, yes, plastic. They can be done as a housed insert, or as a sliding dovetail-shaped insert, however when done in the latter manner repair is more of a hassle if the sill cannot be readily removed for access. Ume-gashi is one term for this treatment; another is mizo-gashi; yet another is ki-suberi.

Once the two infill strips were out, I could see that a simple replacement of the strips was not a possibility as the divider in between was severely worn:

So, I routed it out:

That left a couple of remnants of the center divider at each end, soon to be removed by chisel:


Cleaning up the bottom afterwards:

I fabricated a wenge infill strip, fitted it to the sill, then glued and clamped it down:



The result looked like it would work out well:


Another view:

If this is the original manufacturer’s label, then the tansu was made in Kitayama (on the border between Mie and Nara Prefectures):

Here are the sliding doors. One is original and the other is more recently fabricated, probably on-site, when the building was installed into the Boston Children’s Museum, and has a plexiglass panel fitted. Notice the ebony inserts which have been fitted to the lower tongue on the one door:


A closer look:

The inserts seem kinda cool as a repair at first glance, but after thinking about it for a few moments, you realize that placing a couple of hardwood inserts would only effect, at best, a temporary fix for a worn out rail tongue on a worn out track, and that the hardwood inserts themselves would tend to accelerate wear on the track (since they are harder than the track infill strips) from that point forward. There was very little left of the tongue on with door’s lower rail, and I have simply jointed the edge flat in preparation for affixing a new piece on the bottom.

By fixing old tansu like these, I am learning things which will inform my own furniture designs, so I’m glad to have this opportunity. Wear and tear issues are going to arise with sliding doors in tracks, so what might be good solutions to that…hmm….

All for today, hope to see you again for the next update in this series. On to post IV.

11 Replies to “BCM Tansu Repair (III)”

  1. Thanks for sharing this Chris, studying old furniture like this is a great window to inform future, more durable designs. I can't help but think the product they are getting in return is most certainly “better than new!”

  2. JL,

    nice to receive your comment. I've generally been able to study old furniture by looking at it within the context of museum displays, and it is much more informative to be able to delve into the 'guts' of a piece like this.

    After having spent a great deal of time working through designing a mizuya-dansu myself, it is interesting to come face-to-face with an example to see how they solved the same issues I mulled over. There were certain design aspects which I puzzled over, and now that I see how they were addressed in this particular piece, I come away from it with a rather diminished impression of tansu in general.

    Tansu like these ones seem to manifest, at first glance, a rather utilitarian framed box aesthetic, and one might conclude that their form is largely a manifestation of their function as storage devices. Now that I look into these examples further, I find the form is rather more contrived than that, and I guess I don't identify much with the idea that careful workmanship should only apply to surfaces which are seen.

    I don't really agree with the solutions they came up with for framing now that I have had a chance to look more closely.

    Take the matter of shelf positioning and framing. In a 2012 post called Mizuya (3), I was unsatisfied with the (perceived) limitation of placing shelves only where there were exterior horizontal framing elements which corresponded to she shelf height. On this piece here I see that they placed the shelf where they pleased and simply nailed the shelf ledgers right through the side panels of the cabinet. Not something I would ever do, and I do not admire that they did that, expeditious as it may be.


  3. I wonder where this maker would have fallen in the quality spectrum at the time of production. I would think there are great variations in quality depending on probably both the maker and the client of that maker.

  4. Hi Brian,

    hard to say. I haven't repaired large numbers of tansu by any stretch, but these are similar quality to what I've worked on previously.

    There's quality to be considered in terms of materials, design, and technical execution. Sugi is not, in my estimation, anything other than a run-of-the-mill material, compared to other choices that would have been available. Design has an aesthetic component, of course, but also from a technical perspective relative to the material used. One obvious shortfall I see in this cabinet, and I am by no means sure it is standard practice, is that the boards used for the floor, back and ceiling of the cabinet all run across the short dimension of the cabinet, not the long. This makes no sense to me, unless all they had access to were 2~3' long pieces of wood. Orienting boards across the narrow dimension of the cabinet means the effects of wood movement are going to be maximized, not minimized. And it is surely more work to cut and plane a larger number of small boards than a few longer boards.

    The ceiling boards are cut in random widths, and some are glued to one another and some are not. The framing up top provides convenient points of support at which one can terminate such ceiling boards, however the maker ignored them altogether and just ran the boards across such that the joints between boards fall in empty space and not atop a crosspiece. That makes no sense to me. They tacked on strips to cover the gaps – and I believe this is original to the cabinet – and those strips were held just by nails, which, like the rest, have rusted mostly away.

    Technical execution can be seen in various ways, like the choice of joints, visible tool marks, tolerances of fit, evenness of thickness in planed pieces, and so forth. I haven't found these cabinets too inspiring in that regard. Take for example the sliding door I knocked apart. The lower rail, to which I laminated new front and back pieces, had clear evidence that the groove for the panel above had been originally cut on the wrong side of the stick and was clumsily patched with an infill strip. I would have made a new piece myself if it had been my screw-up. I mean mistakes happen when you make things, but if you have to patch the error, the ideal is to do so sufficient that the error is impossible or difficult to detect. That's a matter of skillful work, not a quality level so much.


  5. That does sound very disappointing.

    It seems very similar to western work in that the top of the pile came out of a limited few shops that were working commissions for royalty and other important types and the middle work was nice looking on the outside by when you dig a little deeper it does disappoint in some ways since they did things to quicken up the process. I see stuff like 3″ over-cuts on dovetails, or the giant boards nailed to the back of the cabinet that were allowed to split….rather than cut a rebate and keep the board intact and flat for years to follow.

  6. Matt,

    it's definitely an interesting switch jumping from the new and precious materials which are hard to work and of which I have little to spare, to a dinged up older cabinet made from a softwood and with insect carcases all over. I'm enjoying the interlude – it's a nice mental break, and while I work on these tansu I am still moving the other cabinet along in a few areas. I've been patinating some hardware, making the central panels for the bonnets, etc.. I've also had another minor job mixed in too, which I have not blogged about.


  7. I have made a few sliding doors, the action has never been quite right it seems, either too tight or a little rattle. Recently while prowling through an antique store I came across an old cabinet with a set of sliding doors. A close look showed the track and the bottom door rail had a very slight bevel, so the door was just slightly snugged to the front edge of the track. Seemed to work nicely.

  8. One reason I can think of to make the boards across the short dimension is to minimize deflection under load. Given that the deflection scales with the third power of the span, it makes sense to make the span as short as possible. This would enable the maker to use thinner boards.

    I was wondering how old this tansu is? It has obviously seen a lot of use, but the label doesn't look that old. But if it was made in the period around WWII, materials and trained labor might have been in short supply.

    Is it usual to apply a lubricant to sliding doors? Before drawer slides were all made of metal or plastic we always kept some candle stubs on hand to rub on wooden sliders. Works pretty well, especially hard paraffin wax (beeswax is on the soft side).

  9. Roland,

    the deflection argument doesn't make sense in the context of a frame and panel cabinet. The panels only bear loads if they are horizontal, and they are supported by frame elements below. Take away the frame elements and the boards, all of 6~8mm thick, would not support much weight. The panels in the back of the cabinet, being vertical, could be employed to bear loads, but they are not, as there are ample gaps around the panel edges at the frame.

    I apply waxy/lubricant (Slip-It) to the sliding tracks and door lower tongues.


Anything to add?

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