Digital Wood Joints(openup.design) |
Digital Wood Joints(openup.design) |
I would caution anyone thinking of doing joinery this way to consider whether it's actually suitable for the application. I was at a coffee shop once where absolutely every chair in the place was starting to fall apart. The reason was that the joinery was all some variation of half-lap, which doesn't constrain the movement of the pieces in all of the directions that matter. Once the glue failed, the chairs started coming apart.
I would also add that the corner joints meant to replace dovetails or a box/beehive joint are unsightly with all the required dogboning and will not improve aesthetically with the addition of glue. I would further point out that there are already quick ways to cut dovetails or box joints with a router quickly and efficiently. It would be hard to convince me that there's a truly useful role for cutting an uglier version of a box joint on a CNC.
Source: am a furniture maker who does some CNC work.
Dowel joints are great, as far as I know Krenov never used anything else on his cabinet carcasses. I'm not sure why they developed a bad reputation.
It looks like everything on the poster is made to be cut on a 3-axis machine. Stepping up to five-axis is a huge leap in cost at present, and surely was then as well.
Tooling has likely improved in availability and cost since 2004 as well. Automatic tool changers have probably also become more affordable, but just like going to five-axis, you're getting into a whole different class of machine once you start talking about adding that.
Some of the reasons to cut stuff single-sided on the flat are unchanged by any of that though. It still costs you precision (and time) to flip a workpiece over, and you're going to have issues if you don't have good consistency with your material thickness if you need to reference your Z axis to the material surface rather than the bed surface.
Working an end of a long piece remains a pain in the ass for fixturing that involves a hole in your machine bed, and possibly the floor as well. Tenoning a bed rail, for instance, is inconvenient however you do it.
All of my CNC experience is on a three-axis machine, five-axis gets you a lot of flexibility that most places won't have unless CNC work is their primary focus, or at least core to their workflow. I've seen a shop that builds high-end windows with a large five axis machine. I have no concerns about the durability of their products, it's just that most people don't have access to that kind of capability.
With the exception of the joints labeled "...with key" these joints are all very remote from the types of joints used in traditional Japaneses temples which do not use glue.
These are mostly western style joints, which are also very beautiful and useful, but generally expected to be assembled with glue.
Great resource!
No. You can't simply use whatever joint you want and expect the glue to deal with the (sometimes enormous) forces applied to it.
I’ve seen YouTube videos where people test different wood joints to see how much load they can withstand, but I assume the results will be different if you account for things like changes in humidity over time, which causes the wood to expand and contract (which doesn’t happen isotropically).
Somebody should run these through a finite-element package to compute their strength. Some of those look fragile.
[1] http://winterdienst.info/50-digital-wood-joints-by-jochen-gr...
Woodwork Joints https://archive.org/details/woodworkjointski0000hayw_k7x4
Hayward discusses not only the joints, but techniques for cutting them with hand tools.
There are plenty of used copies out there. Hayward has many other useful books.
So if you enjoy this, you might enjoy Luke's work too!
[0]https://issuu.com/lukemurrayarc/docs/portfolio_2023 starting at page 26-27.
- Figure out a way to parameterize joints
- Create an automated evaluator that uses finite element analysis
- Evolve joints with a genetic algorithm- constrain joints so that they can be cut with a tool on a CNC with the material secured on the machine in a convenient fashion in a reasonable period of time
which is something I've been working on.
It might be neat to do it automatically based on a picture of the wood, but that is a much bigger problem.
Your interpretation is more likely, though. And explains why so many of the joints are weird and look like they'd leave gaps...
Also, an over abundance of lengthening / scarf type joints, more than the typical woodworker would use, IMO.
They could be made to higher tolerances, and had more surface area, and I assume were more reliable to make/hold draws together.
These are very similar to the idea here.
EDIT: thankyou to mikey_p and WillAdams who pointed out the joint I was looking for was the Knapp Joint: https://www.finewoodworking.com/2018/09/26/history-cove-pin-...
This video has some explanation of the new method and why they were originally popular in the first place (mostly limitations of available machines): https://www.youtube.com/watch?v=MlJjsvph3r8&ab_channel=Matth...
A CNC router affords a lot of new possibilities, e.g., this box which has an integrated lid:
https://community.carbide3d.com/t/as-funny-as-a-3-dollar-box...
To me, these look like a stepping stone to something better. They're all intended to be used with a flat-shaped endmill, and all probably have strength and aesthetic issues. One solution is using different router bit shapes in combination with traditional techniques, but this is more labor-intensive.
1) programming takes a long time, and it only makes sense to take the time to do it if you're making a bunch of copies of something. this is something that could be improved with better software and ux - if cad programs made it easy to just drag and drop joints from a joint library into your model then this would be a different story. a hardware+software solution could also work here, something like a cnc version of https://www.woodpeck.com/multi-router-group.html where the software makes it easy to scale the templates to your work piece.
2) setup takes a long time on the affordable machines. every time you change bits you have to recalibrate. positioning the work piece on the table and clamping/taping it down takes a lot of time. if you have to flip the work piece over then that takes even longer and positioning is even more critical, and programming is more complicated as well. regardless of whether your designs require cutting on one or both sides, you have to program tabs into your design so the router doesn't cut all the way through (or else the piece will move and the router will screw it up), and then you have to go back and cut the pieces out the rest of the way manually and trim off the tabs with a flush trim router bit. the high end production quality machines mitigate a lot of these issues, but now you are talking about a machine that costs at least $100,000 and takes up a whole room.
http://tug.org/TUGboat/tb40-2/tb125adams-3d.pdf
and
https://community.carbide3d.com/t/a-different-sort-of-box/36...
and for more see:
What is really missing for me here is complete pieces, utilizing some or many of these new joints throughout. Are they practical? Do they look good as part of a larger whole? Even 3D renders would be a good start, if not an actual physical piece.
If you want to learn wood joints, learn mortise and tenon, dovetails, dowels or floating tenons, and nailed/screwed joints. Some extra love for biscuit miter joints but really that is all the main ones people use.
https://www.dougukan.jp/exhibition?lang=en
It focuses a lot on the evolution of precise woodworking tools, like saws and planes. They also had examples of complex joints, made without nails or glue.
I can't tell you if that's true or not, but I have a hard time seeing folks putting so much work into developing such joints without a pressing need for them.
And the description for the 3 options is contradictory. Above them it says, "Select a Data Access Level", but selecting "High", the description is "Highest level of privacy"... So what does it do, give the advertisers a high level of access, or give me a high level of privacy?
http://winterdienst.info/50-digital-wood-joints-by-jochen-gr...
As much as I like nice joint design, I think a lot of that aesthetic beauty comes from when it's not machinable (or not traditionally anyway, iirc the pantorouter can make some pretty tight dovetails - because of the orientation vs. trying to route them 'normally'). It's not how fancy and curvey can it look, it's how intricate and fine the detail. Chunky dovetails don't look better than box joints in my opinion.
For a Japanese spin on this see Tsugite:
http://ma-la.com/Tsugite_UIST20.pdf
which I worked through at:
https://community.carbide3d.com/t/a-study-of-joinery/28492
Traditional joints (box, dovetails, or obscure variations such as Knapp (cove and pin)) require a vertical fixture and 3 setups (at a minimum) --- cut parts to length and machine internal features, mount four board and cut joints in 2 corners, flip boards (with correct orientation) and cut other two corners.
Rabbet joints are simpler --- so simple that they were covered in a video as "The Simple Box":
https://www.youtube.com/watch?v=V93xDM3lXsM
(ob. discl., I work for Carbide 3D)
There have been a number of programs developed for joinery. A current commercial option is:
http://www.g-forcecnc.com/jointcam.html
(but it requires a vertical fixture)
One commercial option became freely available:
https://fabrikisto.com/tailmaker-software/
and ingeniously has an option where a 30 degree V endmill is used, but to cut boards held at a 15 degree angle, affording a 90 degree cut with a great deal of control and flexibility --- this can multiply setups to 9.
A variation I've been experimenting with is full-blind box joints:
https://community.carbide3d.com/t/full-blind-box-joints-in-c...
They're reasonably easily drawn up, though they do have some rather specific tooling requirements (a narrow 90 degree V endmill, a square tool of that or smaller diameter, and to make things easier, a large V endmill)
One test project was so tight that after putting it together for a dry-fit before gluing I was unable to get it apart:
https://cutrocket.com/p/63781eaf9822f/
I've been working on a programming system to make this sort of thing a bit easier:
https://github.com/WillAdams/gcodepreview
and have some sketched out joints which I've not been able to make using existing CAM tools which I hope I'll be able to do using this system (if anyone could recommend books on conic sections, I'd be grateful --- that's where I got bogged down last time).
http://winterdienst.info/50-digital-wood-joints-by-jochen-gr...
(which is the original site and which would be better to use/link to)
- design for the worst-case scenario for a given practice setup (note that wood is graded on how many knots there are and how tight they are expected to be) and work around that, eliminating pieces w/ problematic grain - use engineered lumber where appropriate
This is a marked contrast to some styles of hand-woodworking --- I can still recall walking with a great-uncle in the woods when we came to a walnut tree where the trunk had an odd angle to it (apparently the main trunk had broken off vertically in a storm and a branch had continued the upward growth) --- he immediately took a colorful handkerchief out of his pocket and tied it around the tree, looked carefully at a couple of local landmarks visible through the trees, and then confident of being able to return to the spot, we went to his cabin, fetched a saw and ax and proceeded to cut the (smallish) tree down and sectioning it off, taking it back to his cabin where he covered the ends of the central portion around the angle in melted paraffin.
Almost a decade later during another visit, he announced he had something to show me and got out a rifle which I remembered as having a broken stock --- he had of course restocked it (once it had dried sufficiently) with the tree which had grown at an angle which was perfect to make the stock out of.
Sometimes people glue a butt or mitre joint, then once the glue has set, drill a through hole for a through dowel from an outside face of the joint.
The blind holes approach is tricky to get perfect alignment on where dowels are being added across a longer length. As a hobbyist I've tried a few cheaper dowelling jigs and had mixed success. This challenge lead to other loose tenon solutions like biscuits or dominos which allow for some side to side misalignment while retaining the ability to keep the visible faces aligned.
The through dowel approach avoids the misalignment problem, but comes with the visibility of the end grain of the dowel on one exposed face. Some people are ok with that.
Dowels are still one of the "strongest" options for end-grain to long-grain joints. Many professional woodworkers now favour dominos simply due to who quick it allows them to work and the additional allowance for some side-to-side alignment.
A good craftsman would not choose a joint that would see such high stresses.
Additionally if the glue is chosen and applied properly, the wood that the glue adheres to should fail long before the glue.
That said, glue is not as simple as it may seem. There are many different types and proper surface prep and application makes a huge difference to ultimate strength.
For example, many people will mix 2 part epoxy until it "looks mixed" which for a clear epoxy happens pretty quickly. In truth, the resulting bond strength is far more closely related to the amount of mechanical energy that has been transferred into the mixture than the visual uniformity.
Lots of ways to go wrong with glue... but a good craftsman should be well aware of these.
Could one make that a part of the evaluator/fitness function?
i could see how a machine big enough for 4x8 sheets with an automatic tool changer, a vacuum table, and all the automatic calibration gizmos might be a time saver for a production shop, but if you're building something that's a one-off or you don't have all the setup automation goodies (which are $$$$$) then setup and programming usually end up taking longer than doing the work the old fashioned way.
for tenon cutting like in the bed rail example you gave, i have a hard time imagining any situation where cnc is going to be more efficient than a domino xl.
I actually saw an unusually straightforward example of this last year - a group of friends and I were making instances of Tyler Gibson's 1-sheet portable bike rack design (it's great, check it out: https://www.thetylergibson.com/building-a-better-portable-bi... )
One group of two-ish people used jigsaws to manually cut the pieces, and I used a Shopbot 4'x8' CNC router. Very roughly, it took about twice as many man-hours to make one by hand, vs by CNC, and the result was less clean. CNC could have done even better, but due to warping of the sheet, it failed to cut all the way through in places, and I had to do a cleanup pass with the jigsaw. And once the upfront cost of generating the toolpaths etc was paid, it would improve again.
One reputable company has an entire section in their catalog on it: