BCM Tansu Repair (II)

While I had the lower section of the mizuya-dansu laying on its back for purposes of repairing the front sill and attendant sliding track, or shiki-i, I noticed a small piece of wood loose inside the cabinet. I was about to toss it however it looked to be stained on a couple of faces like the rest of the cabinet. So I looked around to see where it might have come from, and lo and behold I found the spot, on the inside lower arris of the left side stile:


Notice how the floor of the mortise cut to receive that piece looks almost like it was hacked out with an axe. Some rough work there, and the piece had not been glued in, but tacked in place with three nails, all of which has rusted to nothing and the piece had fallen loose. You know how much I love nails….

I flipped the cabinet on its end so I could access the area with a trim router, and thereby clean the bottom of the rebate up:


It was looking a little better after that point, though the sidewall was not a flat surface:

I clamped a paring guide in there and cleaned up that sidewall:

Later, I fitted a patch, trimmed it to the surface, and applied some stain to have it match its surroundings.

This cabinet came through customs and into the museum with many insect house remains attached to the underside of the cabinet:

A closer look:

I took the opportunity to remove the unoccupied houses.

Back to the action. After cutting a large housing in the front rail, I fabricated a new shiki-i insert piece out of quartersawn wenge, and glued it into place on the top of the sill:

A minor clamping festival ensued:

After the glued had dried sufficiently, I could get to work on fitting an outrigger piece to help support the track section:

After confirming a good fit, I glued and screwed the piece into place:

As noted in the previous post, I noticed there were some mismatched floor boards in the cabinet. Closer inspection revealed that these panels were Southern Yellow Pine, and had bowed considerably:

Thinking about it later, I concluded that the SYP boards were likely offcuts from the timbers used to repair some of the framing in the house during the installation of the house in the Museum in the 1980’s. That SYP is prone to twisting and warping might not have been something the Japanese carpenters were aware of, and perhaps their material options were limited, along with their time frame for completing the work.

In any case, I thought the SYP panels had to go. Beside an inappropriate material, they were split on their ends from nailing and were warped and bowed. Not much to recommend them really. And they stood out like a sore thumb inside the cabinet. So I knocked them out and pulled the nails.

It was time to refit the front lower rail/sill to the cabinet:

In this view, you can also see the patch completed on that lower left stile, before it had been stained:

The floorboards are done in an odd manner frankly. They are beveled on their ends so as to have a surface which ramps up from the track. I can’t see any sound reason to do this, but whatever.

I had some offcuts of quartersawn Port Orford Cedar which I though would serve well for floor boards:

Note the area on that inner arris of the left stile, the patch disappearing with stain.

The lead edges are beveled to match the neighboring boards and they will be stained to match soon enough:

An errant glob of plaster was found on one side of the cabinet:

With the track back in and ready to roll, I had a look-see at the sliding doors. One was relatively straightforward, needing a bit of clean up of its lower tongue and it was fitting reasonably well. The right side, and front door, however, was uncooperative. I soon discovered what the issue was: the lower rail on the door was badly bowed. So re-cutting the tongue on such a piece was a fool’s errand it seemed to me.

I decided to se if the joinery could be coaxed apart, and found it could:

Clamping the bottom rail allowed me to separate the rest of the frame from the rail tenon:

The bow was about 1/8″ (3mm):

I had decided to make a new rail altogether and then thought of an alternative plan which made more sense to me. I jointed the old rail straight, ran it through the planer, and then made up a couple of laminae to attach, wenge for the front face, and Port Orford Cedar for the rear. One lamina per side makes the construction balanced so it should keep it from further warping.


Another new addition to the cabinet was the lower skirt board, which had rotted and was bug-eaten. The replacement is of wenge, stained to match the rest:

I thought it came out well, which it to say, it doesn’t stand out from its surroundings. The odd thing about the lower skirt boards was their joinery, which is simply that the skirt board ends are fully housed in dadoes. The bottom of the cabinet is tied together with mechanically marginal connections – and there was no trace of glue (though such connections are also poor for integrity even when  glued). It seems the floor boards, nailed in place, are contributing to keeping the bottom of the cabinet tied together (!). Not sure why the maker chose to use such joints, and not sliding dovetails, say, or mortise and tenon, neither of which is significantly more difficult to form than an open slot mortise.

All for this round. Thanks for visiting. Post III is up next.

3 Replies to “BCM Tansu Repair (II)”

  1. Jon,

    thanks for the question. It's a Canadian-made product, SamaN's water based interior stain:


    It does the job in one application, if you can believe it, and does not leave overlapping brush marks or anything of that sort behind. No/low VOC's, and water and soap clean up. No flammability problem. It is premixed, just open the bottle and use it, and you can apparently mix different colors of the product together to fine tune or make custom tints. I like this product! (not that I generally use stain though…)

    I picked the Dark Walnut color, thinking it was a close match to what was on the cabinet, and that turned out to be correct. Lucky guess!


Anything to add?

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