Blind finger joint?

[Jamie Buxton]

A finger joint is a good carcass joint with lots of glue area. For instance, it could be used to join the top to the sides of a solid-lumber bureau, when the grain of the sides goes up/down. However, the usual finger joint leaves the joinery very obvious. I'd like to find a way to make a blind finger joint -- one that has all the structural integrity of the finger joint, but is invisible when assembled. Below is a sketch of what the two pieces would look like. Has anybody seen a jig design which would make this?

[Brad Schmid]

Wow, that's alot of little tiny mortise & tenons :D I can see why you'd want a jig ;) Maybe CNC router with tiny cutter??? Can't wait to see if someone has a simple solution.

On the other hand, how 'bout a different joint, like a stopped sliding dovetail?

Good luck!
Cheers,
Brad

[Jim Becker]

I believe you may be able do this with the Leigh FMT and maybe a variation with the D4 and M2 template system. The tenons and mortises would be rounded, of course.

[Doug Shepard]

Hmmmm. I hope somebody chimes in with a jig idea as I've been contemplating something like a blind mitered finger joint that might be feasible with a jig similar to one that could make your joint. But looking at your sketch, the top piece looks to be the only real challenge. You could get the bottom (vertical) piece with a standard table saw box-joint jig then cut away the shoulders and cut the fingers shorter. Getting that to align correctly with the holes put in the top piece by a separate jig might be pretty tough though.

[Lee DeRaud]

But looking at your sketch, the top piece looks to be the only real challenge. You could get the bottom (vertical) piece with a standard table saw box-joint jig then cut away the shoulders and cut the fingers shorter. Getting that to align correctly with the holes put in the top piece by a separate jig might be pretty tough though.Of course, the solution to that problem is to use the same jig for both pieces and a whole bunch of tiny loose tenons.

The question I have is, is this really going to be all that much stronger than using a couple of loose tenons oriented the "normal" way, or, for that matter, dowels?

[Keith Christopher]

I think what would look even cooler here is if they were thru tenons and sanded flush. I mean what a great symmetrical detail on the top surface. I would think making a jig Ok so here's my idea for a jug to cut the mortises with some "backer blocks" this could also be used to make the tenons.


A long "box" with a threaded screw rod running it length and about 3" past each end. Some cap screws that can be used to tighten the jig to the stock. The sliding frame can move along the surface supported by the "box" a self adhesive measure can run along the support box for accurate repeatable mortises. with an upright support frame and something to back the stock it could also be used to make the tenons. clear as mud ? No ? Well I hope this image helps a little.

Keith

[Jamie Buxton]

Doug -- you're right: the male piece is easy. It is the female piece that is difficult.

Lee -- yeah, dowels sorta work. That's what I do now when I want this kind of joint. I use a template intended for shelf pins and a plunge router. Using those ensures that the dowels line up. With the right template, I get 1/4" dowels on 1" centers. Individual dowels aren't real strong, but having one every inch seems okay. However, always in search of better ways, I like the much larger face grain glue area offered by the joint in my sketch. Of course, if I can't figure out how to build it, the benefit is only theoretical. :(

[Ken Garlock]

Not hard with an Incra router fence system. Make yourself a template for the fence and get with it. :cool:

It is just like making dovetails only different. :eek: Come to think about it, you might be able to use a dovetail template directly. Just change the bit to a straight bit. Put some stop blocks on the fence for the plunge cuts.

[Doug Shepard]

Doug -- you're right: the male piece is easy. It is the female piece that is difficult.


Don't know for sure if this would work well or not, but I had another thought that involves using the male piece as the 'jig' to make the female piece. Make the male piece with a std. TS box joint jig, and make the fingers extra long. Then place the male piece over the female and offset it. Use a bearing guided router bit with the bearing the same size as the cutter, and use the long fingers on the male piece to guide the bit, with an appropriately placed stop block to limit how close to the edge you get. Then trim off the x-long fingers and shoulders on the male piece. This still has plenty of problems - square up the ends of all those holes, or round over all those fingers. I think you'd also have to have finger/slot sizes big enough that you can fit a bearing guided bit between the fingers and the size of the TS cut fingers would have to exactly match the router bit diameter. Just another one of my hare-brained ideas FWIW.

[Steve Cox]

I guess my question would be why? If the tenons went through the piece to be seen on top that would be one thing but otherwise carcase pieces have been assembled for centuries with standard mortise and tenon joinery and I don't see a need for the increase in strength this would give.

[Mike Cutler]

Jamie. I have to ask, what is the purpose of this joint? It would be fun to do I agree, but are you looking for a specific function to be performed.
Depending on the orientation of the grain, I'd fake the joint. I'd cut a standard finger joint and offset the fingers to leave a rabbetted edge. I'd then glue a seperate piece that matched the grain to hide the joint.
The standard finger, or box joint is not inherently very strong by it self. It relys on the strength of the glue, or mechanical fastener for it's strength.
The supers, or boxes, that make up a beehive are normally glued and nailed box joints. A full deep super can weigh 75-90 lbs when full. Even when the joint is epoxied and nailed with ring nails the joints loosen up. Of course they see alot more stress than most rational folks would apply to any joint. 75,000 bees per hive,can really motivate you at times :eek:
Long story made short is, are you looking for greater mechanical strength from the joint you've laid out?

[Mike Parzych]

I'd have to wonder why also, unless you just enjoy that aspect of woodworking. Since modern glues are stronger than the wood they join even in the simplest joinery, I can't imagine an application that demanding of a joint.

[Jamie Buxton]

Ken -- Can you provide a more-detailed explanation about how to use the Incra jig to make this joint?

Steve -- maybe I wasn't clear. The side of this bureau would be planks, edge-glued to form a panel. The top, too, would be a glued-up panel. The grain in the top runs the long direction -- left/right if you're standing in front of it. The grain in the sides runs up/down. The appearance of the carcass is very clean -- if there's no joinery showing. The grain just flows across the top and down the side.

Doug -- using the male fingers as routing guides doesn't seem to me to do the right thing. You want to bore the holes where the fingers are, not where they aren't

Mike Cutler --- I'd say the finger joint is a very strong joint. It has lots of face-grain glue area, and modern glues stick wood together very well.

Mikes Cutler and Parzych -- The purpose of this joint would be to make the kind of carcass I described above.

[Chris Padilla]

Jamie,

The Incra jig gives you precision to 1/32" steps (0.001" repeatability) due to its interlocking design. I have one on my TS and one on my router table...you're welcome to have a go at it any time you like.

The "tenons" or "fingers" would be fairly straight forward although it would involve moving the plank through the cutter and if it is large (as I guess it is), you want a way to move the tool (handheld router) through the wood. Keith's idea of a jig might be helpful in that respect. The Incra stuff is only going to help you if you can move the wood through the tool and not the other way around. However, the Incra could help you make a very precise JIG to use a handheld router with. But, I think you will end up with nice square tenons.

Now the mortises in the other half of the joint could be done on the Incra jig on the router table however they would be rounded and not square.

I'm not sure which is more evil: rounded tenons or rounded mortises.

[Doug Shepard]

...
Doug -- using the male fingers as routing guides doesn't seem to me to do the right thing. You want to bore the holes where the fingers are, not where they aren't
...

That's why I said "Then place the male piece over the female and offset it." Still don't know that this is the best answer, but maybe my earlier description wasn't very clear.

[Mike Parzych]

But do you really need all the tenons illustrated? Wouldn't even one elongated tenon accomplish the task with sufficient strength?

[Ron Taylor]

This application would be as simple as any other dovetail or finger joint on the WoodRat. Stopped or thru would make no difference. Personally, I like the idea of the thru joint.

First you'd do the fingers, same as any other finger joint. Cut the shoulder on the 'Rat, TS, or whatever tool you prefer.

Then run the mortises using the fingers as a direct reference, plunging each mortise individually. But that's where the 'Rat excells....

The one limitation is that the vertical piece, the one with the tenons could be no longer than the height of the WoodRat from the floor. However, this could be circumvented by running all of the tenons on another tool using a typical box joint jig, then make one additional reference set on a short piece for the mortise work on the 'Rat.

And, you'd still have to decide whether to round off the tenons or square the mortises. With a bit of chisel or rasp work, this joint would be very doable on the 'Rat. I'll try and do one soon and post some pics. Sounds like fun :)

[Keith Christopher]

I think that even beyond the strength of this joint it would make an interesting detail on the top if they were thru mortises. Sanding them flush.

[Steve Cox]

Jamie, if the goal is to have clean grain appearance running around the carcase I would submit that there are much easier approaches to this. Blind dados in the top with a rabbet on the side, blind splines, sliding dovetail, even biscuits would give you the appearance you want with enough strength to do the job and you wouldn't be trying to re-invent the wheel. You don't have any wood movement problems to speak of, the grain is going the same direction on both pieces (if you use a spline, orient the grain the same way as the sides/top). I guess it would be fun to come up with a new approach but I'm not a hobbyist and while all my work has to be of excellent quality, I still need to build it as fast and easily as possible.

[John Stevens]

With a bit of chisel or rasp work, this joint would be very doable on the 'Rat. I'll try and do one soon and post some pics. Sounds like fun :)

Ron, I hope you'll be able to do this. Not that I'm interested in this particular joint--I'm more interested in the 'Rat. I'm curious how you'd mount the piece in the rat to cut the mortises. It would have to be perpendicular to the wall the 'Rat was mounted on, no?

[Ron Taylor]

You are correct John, you'd mount the piece face up, perpendicular to the frame of the WoodRat. A bit of care would have to be used here since the work piece would be extended horizontally out from the 'Rat. A shop-built block clamp would be used to secure the piece and some sort of supporting device would have to be used for its length if it were very long. Then, the mortises would be cut with a push-pull motion ("North-South" by the WoodRat jargon) using stop pins to limit the cutter's travel. The WoodRat's feeding mechanism would only be used to position the work piece for the spacing of the mortises. Pics would make it a lot less confusing :cool:

[Jamie Buxton]

Mike Parzych and Steve Cox -- Biscuits and one big long tenon wouldn't be very secure because there would be almost no facegrain-to-facegrain glue area. Glue doesn't stick very well to end-grain.

Chris -- I think rounding each tenon would be evil. Like Brad said, that's a lot of tenons. Actually, I don't worry much about that dimension of a tenon. The important fit is the one where there's face-grain gluing; that's what really holds the joint together. A square-ended tenon in a mortise with rounded ends is structurally okay.

[Ken Garlock]

Ken -- Can you provide a more-detailed explanation about how to use the Incra jig to make this joint?

This is what I would try.

1. Decide the spacing of the tenons. It would be easiest to make the spaces between the tenons equal to the width of the tenons.

2. Take a strip of paper that will fit in the rail of the Incra fence. On that strip mark out the router bit center for the tenons in red. Go back and mark the bit center for the tenon waste removal space in black, they should fall exactly in between the red marks. Make a series of black marks for at least two more than the number of tenons you want to make. Same for the red marks. When you are done, you have a series of red and black marks ending with a black on each end.

3. Decide the inset of the tenons from the edge of the board. Write it down, you will need it later.

4. Set the router exposure to the value from #3 above. Run both sides of the tenon board thru the router with the soon to be tenons pointing down.

5. Attach the 90 degree jig to the fence, and clamp the tenon board to it vertically with the soon to be tenons pointing down and the board at right angle to the fence.

6. Calibrate the board to the template you made above. Easiest way is to match the router bit to the center of the board and move the center tenon mark on the template to match.

7. Using the black marks, cut the tenons on the end of the board by removing the waste in between them.

Making the mortises:

8. Take the inset value from #3 above and place a stop block on the fence "down stream" from the bit trailing edge.

9. Measure the width of the tenons. Higher math: measure the board length, subtract the #3 value, and then subtract the tenon size( it should be board thickness minus two #3 values.) The resulting number is the place for the second stop block up stream from the bit leading edge.

10. Cut the mortices by using the red marks on the template AND positioning the board at the up stream stop block. Do a plunge cut into the board and route up to the second stop block down stream.

11. Each and every cut will require a resetting of the fence to the next red or black mark on the template.

This is very similar to how you cut dovetails on the Incra system, except they supply the templates :)

It makes sense to me, I hope it does to you also.

[Jamie Buxton]

This has been an interesting thread for me. Thank you all for your suggestions. As I've thought about this problem yesterday and today, I've come up with two solutions for the case where the inside of the joint is hidden -- as in the bureau I mentioned.

One solution is pocket screws! :p After I thought of that, I was surprised that one of you Kreg-jig fans hadn't suggested it.

Another solution is what I'd call a blind mitered finger joint using loose tenons. (How's that for fancy?) It has the strength of the finger joint, while letting the grain flow over the joint. It also can be made pretty easily with a router.

The first pic shows one side of the joint (say, the top of the bureau) from the inside face. It is red. The blue piece floating above it is the router jig which is used to cut all the slots. The jig is made with a dado head, so that the slots in it are all exactly the same width; this will be the thickness of the loose tenons. The spacing between the slots doesn't matter. This is good, because I can make the jig without careful measurement or construction -- I just slam some slots through the jig board. The jig is first used to machine the slots in the first workpiece, and then flipped over to make similar slots in the mating workpiece. When it is flipped correctly, the slots in the two workpieces line up, regardless of any irregularity of the slots' spacing.

Another good thing about this jig is that it doesn't require plunge routing. The router just rides in from the end of the jig.

The second drawing shows most of the loose tenons inserted in the slots, with one as yet uninserted. It also shows the two sides of the joint, almost ready to be assembled.

[Chris Padilla]

Very nice, Jamie! I think you got us all stuck in a "rut" with your first pic so no one ventured outside of that.

I've seen the loose tenon joint before only the loose tenons where much larger...nearly perfect cubes if I remember right. You could probably cut the number of mortises/tenons in half or one-third by making a fatter tenon if you wish. Also, you could completely hide the mortise/tenon...again, if you wish.

At this point, why not some biscuits? You'd be done in 1% of the time me thinks! :)

[Jamie Buxton]

.....At this point, why not some biscuits? You'd be done in 1% of the time me thinks! :)

Unless you do something very tricky with the biscuits, the biscuits' grain would run in the wrong direction. In one side of the joint or the other, there would be no facegrain-to-facegrain glue area.

[Mike Parzych]

Well.....have fun.

[Ken Garlock]

Here comes Jamie to the forum and asks how to do hidden box joints, a reasonable request, right? :) So, everybody runs off and devises their solution and posts it, some just comment on the idea.

Then we find out that everyone was wrong, and the real solution was mini-biscuits on their side. Hummmm, biscuits on their side, hey, we should call them dumplings and then we can start selling dumpling slot makers ;) :D

You know, Jamie, I think I had a teacher like you once :rolleyes: :D :D :eek: Anyway, it was a fun exercise :) :)

The above has not been approved by the humor control board, and the content is the fault of the author. :rolleyes:

[Chris Padilla]

Unless you do something very tricky with the biscuits, the biscuits' grain would run in the wrong direction. In one side of the joint or the other, there would be no facegrain-to-facegrain glue area.

I think Ken was alluding to that very idea in his rather humorous post: Make your own biscuits with the grain running appropriately or chop up the factory ones.... :rolleyes:

[Richard Wolf]

Does anyone else feel like the wheel was just reinvented!!

Richard

[Mike Parzych]

Quite frankly, I've never seen such a monumentally over-engineered solution to a rather simple task.

[Keith Christopher]

Funny thing is this is more a spline joint than a finger joint :eek:

[Dev Emch]

This can be done with ease using an oscillating chisel mortiser like a Maka STV-160 which is no longer being made. A similar machine is being made by Lori and Lori and one of these is even CNC but that is not needed.

The Maka can blow the square holes into the wood with the speed and accuracy of a router. You really need to see how one works to understand it. The geometry is like nothing you have ever seen.

[Richard Wolf]

Dev, You definitely have a knack for simplifying things.


Richard

[Mike Cutler]

Jamie. One other joint that you may want to consider is a mitered, 45 deg. half blind dovetail. In your original post I thought for some reason that you wanted to keep the Butt joint appearance.
I still think that you are relying too much on the adhesive componenet in the joint, and not enough in the inherent strength of a good mechanical joint. However, that's a personal opinion, and it is not my project, ergo it's not my call. No matter what though, have fun and let us see the final result. It's an interesting concept, personal bias aside. ;)

[Mike Parzych]

On the multiple vs. one long tenon issue:


Mike Parzych and Steve Cox -- Biscuits and one big long tenon wouldn't be very secure because there would be almost no facegrain-to-facegrain glue area. Glue doesn't stick very well to end-grain.


Figure it out. Say for instance you consider 10 tenons with 1 square inch per side, spaced an inch apart...4 sq inch per tenon times 10 = 40 square inches of face/face grain contact over 19 total inches.

As opposed to one long tenon 1" high by 19" long, and by golly, that equals 40 square inches of face/face grain contact. Plus you've got even more end grain contact.

[lou sansone]

Does anyone else feel like the wheel was just reinvented!!

Richard

I agree with richard on this one! I see no purpose in it since the joint is hidden. but hey .... have fun
lou

[Dev Emch]

I agree with both Richard and Lou....

Personally, I would not embark on this excercise; however, should I, the Maka will make child's play out of the mortises.

You all know how to cut the tenons using a simple jig on a table saw with a dado setup. You agree that that that is child's play?

Using darn near the same simple jig on the Maka, you can blow these square holes in no time flat.

I use the maka on face frames. That is my "secret weapon" machine. I got one of the last three imported prior to Maka stoping them. Found it by accident in an automotive repo garage of all places. No one knew what it was so I snagged it for a song. Like I said, the geometry is so wild and out there that even woodworkers could not figure out what the heck this thing actually did!

Speed, accuracy and uniqueness is how you make the real money. The Maka gives me that. Its only drawback is that it cannot make a SQUARE hole smaller than 3/4 by 3/4 inch. Mortises that are say 1/4 in by 3/8 in or even the length of the board are no problem. Perfect square haunch mortises are cake. Stickley spindles that are 1/4 in by 1/4 or 1/2 in by 1/2 are a problem and that is why I got the oliver.

So you guys wanted a quick solution... here it is. This can be done but why would you want to? One final thought. Your going to quite an extreme to complete this joinery yet it is not a locking joint. Personally, I would much rather have a set of dovetails holding this whole thing togther or just use through fingers.

[Brad Schmid]

Hey Dev,

Could I pursuade you to post a picture of this Maka machine? Maybe I've been livin' under the rock too long, but I'll be the first to admit i've never seen nor heard of it :o

Thanks,
Brad

[Daniel Rabinovitz]

Forget all those loose tenons.
Just make it a 45 degree on both pieces, one with 1/4 inch tongues, and the other with 1/4 inch slots.
Cut the slot board to a 45 first on the table saw - then make the slots
Set the slot board on a 45 and run the cutter straight in; from the inside face side.

Set the tongue board on a 45 and run the cutter in; from the inside face side.
Offset 1/4 inch cut
Tough to describe but I think I got it in MY head.

[Jim Becker]

Dan good idea, but I suspect he'd want them to be stopped cuts so they don't show. Biscuits as splines would accomplish the same thing, in a sense.

[Jamie Buxton]

I tried the method I outlined above to make a blind finger joint. It worked very well, and was rather easy. In an afternoon, I made the router template, cut the joints, made the loose fingers, and glued together the casework for three bureaus. In the first photo, you can see a panel with the loose fingers glued in, and the router jig above it. In the second photo, you can see the outside of the joint after glue-up. As you can see, there is a nice clean appearance as the grain just flows over the edge, with no endgrain visible. If you want, you can make the planks run continuously as they go over the edge. I didn't do that here because the three bureaus will stand side by side by side, so I made the grain run continously across all three tops.

[Jim Becker]

Excellent, Jamie! That's a nice, clean, contemporary look. If I had not already assembled the cabinet for my bench, I would have liked to try this out since I did the "wrap" thing with the plywood running continuous from side to top to side on that case.

I like how you indicated you ran the grain across the tops since the units will be nested side-by-side. Good forethought.

[Roy Wall]

very nice Jamie..........that's one heck of a joint!!

[Chris Padilla]

More pics, more pics!! :) That turned out extremely well, Jamie, although I'm not surprised with you at the helm! Very clean, very clean.

[Sam Blasco]

If you are not trying to show off your skills and the joinery, a good lock-miter joint would be very quick, very strong, and so-so easy to glue. It would offer the same look as Jamie's piece, and it is a nice look. I use this joint all the time. You just need a shaper or router table and the appropriate cutter (many manufacturers - and in some cases 2 matched cutters.) You could practically do 20 as fast as you could do 1. But if you want to have fun with an array of different tools and jigs and layout devices -- have at it and enjoy the experience to your favorite tunes. The quickest road is often not the most scenic.

[Chris Padilla]

Good point, Sam! I had totally forgotten that joint. I would have worked quite well here. I haven't had very good luck with it on plywood as it tends to blow out the plys on one piece and do very nicely on the other piece--maybe my plywood ain't that great!?

I have also found it, for me, to be very difficult to set up. Now I'm a well-edumacated man but I'll be darned if I can set this bit up without about 10 test cuts and I've read several dittys on how to "easily" do it! Confounded bit! :)

[Jamie Buxton]

If you are not trying to show off your skills and the joinery, a good lock-miter joint would be very quick, very strong, and so-so easy to glue. It would offer the same look as Jamie's piece, and it is a nice look. I use this joint all the time. You just need a shaper or router table and the appropriate cutter (many manufacturers - and in some cases 2 matched cutters.) You could practically do 20 as fast as you could do 1. But if you want to have fun with an array of different tools and jigs and layout devices -- have at it and enjoy the experience to your favorite tunes. The quickest road is often not the most scenic.

It is true that a lock miter would be much faster to make. However, I've steered clear of it when the grain orientations are as shown in the photo. For every part of the glue line, at least one face would be end-grain. End-grain glue joints aren't very strong.

[Robert Ziegler]

I would have to agree with the Mikes - why would you want to use such a joint? Floating tenons would work just as well. Regular mortises and tenons would do just fine - maybe two or three depending on the size/width.

The joint you're making isn't visible. Modern glues will handle any joinery strength issue.

As to your answer to the Mikes as to their question of why? Your response To make the carcass described above isn't an answer. So, I ask again - why?

If you felt you absolutely had to do finger joints, I would do as described by others - do regular finger joints and rabbett and cover them after assembly.

Looks like overkill to me.

[Jamie Buxton]

I would have to agree with the Mikes - why would you want to use such a joint? Floating tenons would work just as well. Regular mortises and tenons would do just fine - maybe two or three depending on the size/width.


Thiink about the grain directions in the two pieces being joined. Big wide mortises would mostly have glue lines in which one face is end-grain. Glue doesn't stick well to end grain. Only at the ends of the mortises would there be face-grain to face-grain contact. That's the contact you want for a strong glue bond. The finger joint has oodles of these face-grain to face-grain areas, which is what makes it such a strong joint.

[Robert Ziegler]

Thiink about the grain directions in the two pieces being joined. Big wide mortises would mostly have glue lines in which one face is end-grain. Glue doesn't stick well to end grain. Only at the ends of the mortises would there be face-grain to face-grain contact. That's the contact you want for a strong glue bond. The finger joint has oodles of these face-grain to face-grain areas, which is what makes it such a strong joint.

Okay, so the only difference between my two or three tenons and the finger joints in the original question is the number of tenons/fingers providing face-face grain surfaces. I understand your point of face-grain to end-grain being a weak joint. If this were a butt joint, I would agree. But the mortice and tenon provide their own strength made stronger by the glue. You could put a thru tenon in this piece, wedge it, and use no glue at all and I would wager it would hold up. I remain convinced this is overkill unless there's a real good reason to do so.
But hey, this is just me talking here. Others have figured out how to do this joint. Okay by me. Just scratching my head politely.
I mean it is an interesting woodworking conundrum - how to do that joint - and I will add the techniques to my files for future reference. :)

[Jim Becker]

Robert, I think that Jamie is trying to accomplish a totally blind joint...and if you look at the pics he posted earlier, you'll see the result. There is no joinery showing. Using another form of visible joinery, rebating and then covering would result in a totally visible situation.

[Robert Ziegler]

OkieDokie - looked at the mitered finger joint photos. Looks great. Rock on.

[Doug Shepard]

I like what you came up with for plywood, and despite some of the flak you're taking I'm slightly disappointed that you didn't pursue your original quest of doing a true blind fingerjoint. For solid wood, I would think that might be a little stronger joint than all the floating tenons. Given what you came up with here, I would have liked to see what you came up with for the original joint if you'd stuck with it.
Regardless - it's a nice solution.