A Ming-Inspired Cabinet (54)

Today was the day: glue up of cabinet 1. This is where months of work can be ruined in less than an hour, so I was naturally a bit on edge.

Prior to gluing, I masked off where required, and used a gouge to make a pair of slight hollows on the side of each dovetail pin:

I did the same on the sides of the tails. The dovetail joints are a tight fit, so the hollows ensure that there will be some glue trapped within and it won’t nearly all get squeezed out by assembly. When I use glue, I try to do so judiciously, not just slathering it on everywhere. My goal is an absolute minimum of squeeze out, and I only apply glue to places where the grain-to-grain bond is a sound one. I keep glue off of end grain insofar as possible.

I was doing this glue-up solo, and therefore needed the maximum glue open time I could obtain. This ruled out hide glue and aliphatic (PVA) glue. Resorcinol would have been an option, however shop temperature is below 70˚ to that was out. I opted for a System 3 epoxy product, T-88:

This 2-part adhesive gives 45~60 minutes of open time, and cures slowly. Slow-curing epoxy is the strongest epoxy, but I mostly was after the long open time. As it turned out, the shop temp of 55˚ or so made the set slower yet, so a full hour proved to be the working time. The drawback to this epoxy is that the clamps need to be left on overnight and a full 24 hours is required for the initial cure, with the cure continuing for a further couple of days beyond that.

Still, when the epoxy was mixed, I was uncertain how much time I had or what might eventuate, so it was full steam ahead. I filmed the whole thing, but after I was done I discovered that the camera position was such that the top of the cabinet was out of view. That was too bad, but “learn as you go”. So, I’ll do another video, with a better camera angle, when I put cabinet #2 together in the near future.

When the dust had settled, things had gone totally as planned, the cabinet was together and looking good:

I took a walk outside and sat by the bank of the river for a while to relax.

I was of course wanting all the joints which were just connected, some 32 mortise and tenons, plus the dovetails, to be drawn up nice and tight with no gaps. And that was the outcome – whew!

Here’s one junction:


Another one:



Another rail-meets-carcase side:

A few photos now to show some of the longer joint interfaces, where the fits came out as I wanted and a minimum of glue squeeze out. This is the junction of the upper shelf panel and rail to the carcase side:

The inside of the upper carcase corner:

One drop of squeeze out along that junction is definitely something I can live with.

The opposite upper corner – again, one drop of squeeze-out:

Shelf and rail meets right side carcase wall:

I managed to stand the beast up – with the clamps attached it was especially heavy:

Another view:

The back side:


As the day drew to a close, I had the upper shelf assembly glued and joined onto the second cabinet:

I intend to be slightly more organized for the next glue up – lessons learned – and hope to produce a video to boot. Stay tuned for more, and thanks for visiting. Post 55 is up next.

17 Replies to “A Ming-Inspired Cabinet (54)”

  1. Chris,

    Looking good! I'm glad that the stress of a glue-up was well spent…I can't count the number of “Oh shit!” moments I've had once the piece was full of PVA. I can't begin to express the gratitude of having a process like this- and of other projects- documented in such a way that shows the actual process of building a piece, not just the highlights but the ups and downs of the learning process that continues throughout one's journey toward being daiku. Such has been my journey, and your steps ahead of my own have served as inspiration to keep plodding along.



  2. Ralph,

    I liked the fact that epoxy is intended for structural use, intended to extreme conditions, cold conditions, submersion, etc. If you look on the back of PVA glue bottles, Titebond, etc., you'll see that they say that the product is not intended for structural use, and not recommended below 60˚F, etc.. Epoxy can also be readily tinted, which opens up other possibilities: T-88 dries clear if untinted.

    The last time I glued up a large dovetailed box (around 2005), also in bubinga, I used some type of resorcinol glue, which worked well and has a decent open time, however it's meant for use at 70˚F or above.

    Thanks for your comment.


  3. Chris,

    is the T-88 the same glue you used to fill the cracks (end grain on main posts) during the MFA gate build?

    I love the way how you present your build threads – little by little, piece by piece, step by step – the reader has to fiddle about the joinery and how it will all go together. And sometimes you solve a part of the puzzle and one can see if his/her assumptions/conclusions were correct.

    Locking forward to the rest of your joinery solutions for this project.

    I said allready that this will be a great piece of furniture?


  4. Mmhh, one question about the drawer stops.
    In post 51 you mentioned that you “still have to mortice the lower carcase boards for the stops”.
    Do you opt for another plan?


  5. Marc,

    yes, I picked up the T-88 for the gate build. You have a good memory.

    I'm happy that you like the style of this blog when doing a build thread. Possibly I could include drawings a little more often to clarify which piece of the project I am working on in an installment.


  6. Potomacker,

    I guess you're razzing me. Fair enough I guess.

    Yes, I did consider that, and of course there is no reversibility. It would have been nice to use hide glue, but it just sets up too fast to make it a good choice. For this build a dovetail carcase made a lot of sense on many fronts, but one of the trade offs is the fact that the joinery cannot be taken apart again. It's the main reason I normally avoid using carcase-joined construction.


  7. In a way, epoxy *is* reversible. If you heat it over the glass transition temperature (“Tg”) it softens (like any thermoset). The datasheet for T-88 doesn't specify the Tg, but it lists the heat deflection temperature as 119 °F and the maximum service temperature as 180 °F (≈80 °C).

    At a guess, once you get into the 100-120 °C range (≈200-250 °F) the glue would soften enough to pry loose the connection. This is assuming the wood holds at this temperature, of course. But I've seen wooden pallets go into 250 °F ovens without apparent ill effects, so I assume that would be OK if handled carefully.

    I've used this trick to disassemble aluminium parts that were bonded out-of-tolerance to G-11 fiberglass strips.

  8. Don't get me wrong. I think you give enough informations to be clear on which piece you are working.
    Rather I wanted to express that I personally enjoy the examination of the subject.
    It never hurts when you have to make some own thoughts.

    My favorite example: The hammerhead corner connection of the “Square Deal” tables.


  9. Marc,

    I appreciate the return, and I like the way you think about things. Filling in the gaps can be good, so long as we're not talking actual joinery!


  10. It wasn't my intent to razz you as you young folks like to say. But the choice of epoxy did seem to undo so much of the efforts that had gone into the joinery and ensure that this piece will outlast us all and more.
    I'm certainly aware now of the Tg idea and am going to look into its application.

  11. Mitchell,

    I don't follow you – how does the choice to use adhesive affect the integrity of the joinery or the durability of the piece? What efforts, joinery-wise, does it undo? I don't think the fit of the joinery could have been significantly better than it was. The only thing it takes away, it seems to me, is future ready disassembly of the main cabinet. Instead, the cabinet breaks down into larger chunks (stand, body, and top). The stand can be completely disassembled. I thought that this was a reasonable trade-off overall.

    As I think you are aware, I do seek to employ a minimum amount of glue and fasteners, preferring the virtue of demountable joined connections, but that does not mean I am a slave to the idea or rigid about that rule.

    I could have come up with some sort of joinery variation for the carcase which made use of no glue, however in design there are trade-offs which I must weigh carefully. In this case, I wanted the joinery to be somewhat quieter, and any other forms of un-glued connections at the box corners or between shelf framing and box, at least as far as what I can envision, would have been far 'louder', so the choice was clear to me.

    Ideals and philosophy of approach do guide design, but they should not exert strict control over it. I made the choice in the past to push the envelope by designing pieces to use joinery only, and this has been, IME, a great way to develop my joinery skill. At the end of the day, however, joinery is not the sole factor used in deciding how to construct a cabinet.

Anything to add?

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