Over the past weekend my wife and I took the opportunity to make another trip down to Springfield MA and visit a museum there. They had a new exhibit on featuring curios and curio cabinets, which included a small selection of tansu. That exhibit was a bit underwhelming however there were lots of neat things to look at besides in other sections of the museum. One of those was a Japanese samurai armor cabinet:
While I had noticed this piece in the past, this time I was struck by the fact that it had bifold doors, like my cabinet design. Bifold doors are not terribly common in Japanese or Chinese furniture, but they do exist. This particular cabinet is larger than most others I have seen- not that it is an especially common form of cabinet in the first place.
On my design of cabinet, I am doing an unconventional form of frame and panel assembly for the door leaves. This method makes the doors a bit thicker overall than normal; otherwise I am concealing the frames behind the panels. With tall skinny doors, a conventional frame and panel has a slight drawback in the amount of room, visually, that the frames must take up at a minimum. The part often to be emphasized is after all the panel itself, a design direction which tends to push the frame members to be as slender a section as is practical -and as may be accommodated with suitably strong joinery. With the frame members becoming increasingly slender so as to make the panel more prominent, the risk increases that the frame corner joins, typically glued, will eventually give slightly, causing the door leaf to distort. This effect will be more pronounced with a bifold door. Make the frame thicker to afford stronger connections and the view of the panels shrinks commensurately.
Such distortion from shear has occurred with this cabinet’s door corner joints at the museum – the inner leafs have sagged relative to the outer ones:
This situation could be ameliorated significantly if several battens were affixed to the rear of the panel to keep it flat, and those battens are then also tenoned into the frame. I strongly suspect that the doors in the above pictured cabinet do not have battens on the backside, at least not ones which are mechanically connected to both the panel and frame. Some people consider the main reason to employ dovetailed battens is for keeping the panels flat over time, which they certainly do, however the other function is to greatly stiffen up the entire door assembly against shear loads as are imposed by gravity. If you only consider the first function, then on a slender door one might conclude that battens are unnecessary since the panels are quite narrow. This is a mistake I think.
These bifold doors are hinged on the front and side surfaces, which means the doors can only open 180˚ and if the door is bumped when in the fully open position then the hinges take a lot of strain from the leverage. This view from the side also reveals no evidence of through mortises for any door battens, as you would typically see on a classic Chinese cabinet:
With a thin frame, through mortises are going to give the maximum strength at the connections, as compared to blind tenons.
The cumulative sag in the door frame assemblies is quite clear to see in this view of the lower portion:
Also, the type of front door latch does little to keep the doors themselves flat and tight to the surrounding casework. Below the doors are a pair of drawers with frame and panel fronts. I am thinking it likely that my glue-less drawer design could also be employed with frame and panel drawer fronts, and that is something I will explore with a drawing when I get a chance.
The door panels themselves did not have the cleanest surfaces:
This view shows one of the blemishes a bit more clearly:
Not sure if that blemish – well, let’s call it what it is, a crater – or the many other ones to be found in the panels, occurred after manufacture or during. It’s part of the story of the piece.
Anyway, it was educational to look at that 100-year old cabinet, as it confirmed some design-related suppositions I have had in recent months, and I now feel confident that I am on the right track with my bifold door design. Time will tell.
Back at the shop….
Currently working on the joinery for the drawer horizontal dividers, here processing the sliding dovetail mortises for the drawer runners:
Obviously, these are not ‘typical’ framing joints for web frames, but rather a result of problem solving when one desires to avoid the use of glue.
One by one, you get ’em done:
Another view shows the tongue and groove portions of these two-direction rod joints, nihō sao shachi sen tsugi:
A version of this joint is detailed in tAJCD Volume III.
This joint slides together as follows:
Another view:
The drawer runner stock lay below.
Each of these sticks is actually composed of three members:
A side view reveals the connections:
The sword tip miters remain to be cut so the joint cannot fully close up as of yet.
Another view:
Another view:
When the joints are drawn up, the spaces now seen at the end of the tenons will be considerably smaller, but not zero.
From the back a portion of one of the internal tongue and grooves can be seen:
The tongue and groove forms an enhancement to this connection, a change I made for some good reasons.
Dry fitted, these connections were plenty rigid and tight enough that I could super-surface the assemblies, which saved timer and made all the surfaces co-planar:
The super surfacing removed all traces of tear out from the VG bubinga, not an easy wood to wrangle.
Also worked today were the horizontal drawer dividers used on the back side of the cabinets:
These, as you can see, are one-piece assemblies rather than three-piece. The connections used to attach these to the vertical dividers are simple half laps:
All the drawer runners, 24 pieces, have been dimensioned with high precision and will have their joinery completed next time I’m back to the shop. Then it will be on to the drawer vertical dividers, which have been planed close to finish thickness. Things are rolling along.
All for today, thanks for dropping by the Carpentry Way. On to post 37
not to clear as to how one could tenon battens to the side rails and have them fixed to the back of the panels, would they have sliding dovetail attachment to the panel giving them greater depth to allow for a tenon, it seems the distance between the panel and the face of the rails is quite small or am i missing somethingi agree that a through tenon at the corners would be stronger esp if if had a peg through it.
the making of this cabinet is an interesting process and quite complicated, am i correct that no glue is used in the construction
do you have any thoughts on non animal japanese wood glue, rice glue etc and where to procure some
thanks for the postings really enjoying it/
not to clear as to how one could tenon battens to the side rails and have them fixed to the back of the panels, would they have sliding dovetail attachment to the panel giving them greater depth to allow for a tenon, it seems the distance between the panel and the face of the rails is quite small or am i missing somethingi agree that a through tenon at the corners would be stronger esp if if had a peg through it.
the making of this cabinet is an interesting process and quite complicated, am i correct that no glue is used in the construction
do you have any thoughts on non animal japanese wood glue, rice glue etc and where to procure some
thanks for the postings really enjoying it/
Plato,
thanks for venturing a comment. In regards to cabinet door construction, a picture would have certainly been worth a thousand words there huh? The battens have a sliding dovetail, which typically is going to enter about halfway into the thickness of the panel. The tenon on the end of the batten therefore is flush to that surface (the bottom of the dovetail trench). The panel is generally going to be relatively thin and could be pushed forward in the surrounding frame to give more space for battens at the back, while the panel's tongue fits into about the middle portion of the stile and rail sections.
If there is adequate depth on the back of the cabinet for a batten to be placed, then there will certainly be room for a tenon on the end of that batten. If the batten is not dovetailed into the panel however, then there would be no room for a batten tenon unless the framework were made deeper.
My aim in joined pieces is to use a minimum of glue – to rely upon it as little as possible. I also appreciate demountability as it makes things easily repairable and transportable. These are virtues of Chinese classical furniture. So that is the aim, however, sometimes the design direction in other respects will make some glue use inevitable. This cabinet is a case in point. The support stand will have no glue at all. The main carcase is a dovetailed box so I will be gluing the corners together, and since that traps the framing within, some of the framing members may as well be glued too. Again though, I seek to rely upon joinery for long term integrity, not glue. So, where a tenon is used that cannot be pegged, I will wedge it and apply a dab of glue only to the wedges.
In my recent adventures resawing stock, I had one shortfall in material for a back panel, so I had to glue up one panel from two pieces. I regretted that, but can live with it.
Glues have various bonding strengths. If I use glue on a piece which I never want to come apart, like that back panel made from two boards, I use epoxy. For gluing the carcase dovetails I will use hide glue, which, as I'm sure you know, is reversible. Rice paste glue is a weaker bond than the fish or animal hoof glues, however it is fine for providing some reinforcement to a mortise and tenon joint that does not permit, for one reason or another, conventional mechanical pegging or wedging. Rice paste glue is easy to make from rice, so it's something you make yourself rather than sourcing commercially. One could perhaps substitute another starch based glue for the rice paste, such as wallpaper glue. Chinese furniture makers of the past often swabbed joints with (natural) lacquer before assembly, so there is another option.
Appreciate that you noted that you are enjoying this series so far.
~C
I'm really enjoying this build Chris, and I'm excited to see how the horizontal drawer dividers assemble to the rest of the cabinet. The double rod tenon assemblies are really amazing pieces of joinery!
– Jon
Jon,
thanks, I appreciate that. Glad you're enjoying the build so far.
~C