Mortise and tenon question

[Tom Porter]

I’m planning a workbench with mortise and tenon joinery, with which I have very limited experience. I’ve been looking into tutorials on the various ways to cut the mortise. In one instance, the instructor cut them with a router but did not square the round corners with a chisel. He considered the step to be unnecessary, reasoning that there was sufficient surface area for the glue to make the joint strong. This doesn’t seem to be the prevailing opinion and seems dicey to me. I’m curious as to whether this practice is common or a bad idea.

I’m not particularly considering making the mortises this way; I’m just curious.

[Brian Holcombe]

Mark out the mortise to match the tenon (or vice versa) and cut to exact size. A router works for big mortises, I keep away from the knife lines then cut to them with a chisel.

If you make through mortises they’re easier and stronger because you can cut from both sides toward center.

[Simon MacGowen]

He considered the step to be unnecessary, reasoning that there was sufficient surface area for the glue to make the joint strong.

He is right from a structural point of view. Those ends are endgrain to facegrain, and the critical part is the facegrain to the mortise wall.

Festool dominoes (loose mortise and tenon) have curved/pointed ends, not square, and again, the critical glue surfaces are the faces and the walls.

Squaring the routed mortises is not difficult (it is good chisel exercise), or you can round the tenons to match the mortises. Or, just leave them alone as the instructor suggested.

Simon

[Chuck Nickerson]

Lots of people make mortise and tenon joints with rounded corners. After all that's the Festool Domino is. Dowels are an abbreviated version al well.

FWIW- I 'm a square-corner guy myself.

[David Eisenhauer]

While the proponent of the non-squared corner mortises may be correct in that there is sufficient glue surface such that it is not necessary to square the corners, I do not believe that practise is widely used. Large mortises (such as in workbench base construction) do often have the bulk of their material removed by a router or drilling, but the final sizing to the line is accomplished by a chisel as Brian says.

[John TenEyck]

I've been using routed mortises and loose tenons for over 20 years and have yet to have a joint failure. That's proof enough to me that they work. Aesthetics might dictate the use of square mortises on a through joint. Otherwise, I favor routed ones with loose tenons for their speed, simplicity of process, and well mated surfaces.

John

[Andrew Seemann]

I don't know that I would use a rounded mortice with a square edged tenon. Rounded mortices with rounded tenons work just fine. Personally I like the idea that there being some physical bearing surface on the end of the tenon, even if it isn't as good of a glue joint. Theoretically it probably has enough glue surface on the sides, but theory doesn't always work well in actual woodworking practice.

Myself I always use square tenons and mortices. Mostly because my mortiser is much quieter than my router.

[Simon MacGowen]

I've been using routed mortises and loose tenons for over 20 years and have yet to have a joint failure. That's proof enough to me that they work. Aesthetics might dictate the use of square mortises on a through joint. Otherwise, I favor routed ones with loose tenons for their speed, simplicity of process, and well mated surfaces.

John

Frankly, joint failures are less common these days because of the modern glues we use. They do happen, but a lot because of poor execution (filling gap with glue as a fix to an undersized tenon or tail for example). I have seen only one domino joint breaking apart due to overleading and undersized members, not the fault of the dominoes (or their non-squared ends).

I still use the router and loose M&T but only when the job is too big for the dominoes.

Simon

[Pat Barry]

A mortice with rounded ends, ie drilled with a Forstner bit and a matching tenon is just as strong as one that is squared off with a chisel. A good fit and adequate adhesive are primary concerns.

[David Eisenhauer]

Agreed. I was under the impression that the OP meant that a rounded corner mortise (created by a router) could be mated to a rectangular cornered tenon because the sidewall glue up of the tenon to the mortise alone was of sufficient strength to negate the need for a snug square to square matchup (assume it also holds for a round to round matchup). As I guessed above, probably the glue up would be strong enough, but I just don't think many folks are going that route. I bet that most folks match their square to square and round to round corners for a snug fit.

[John TenEyck]

Agreed. I was under the impression that the OP meant that a rounded corner mortise (created by a router) could be mated to a rectangular cornered tenon because the sidewall glue up of the tenon to the mortise alone was of sufficient strength to negate the need for a snug square to square matchup (assume it also holds for a round to round matchup). As I guessed above, probably the glue up would be strong enough, but I just don't think many folks are going that route. I bet that most folks match their square to square and round to round corners for a snug fit.


I intentionally make the tenon at least 1/16" narrower than the mortise so that I have some room for adjustment to align the parts during glue-up, in case I didn't get the mortises perfectly matched. The glued side walls are what hold the joint together.

John

[Brian Holcombe]

They hold the joint together but the end grain abutments are important to the structure. I don’t agree that an intentionally loose fit at the ends is the same structurally as a tightly fitted tenon.

[Simon MacGowen]

The round mortise-square tenon is structurally strong for regular furniture projects. As I pointed out in my last post, if such a round mortise-square tenon is going to fail, it is most likely due to something else (loose fit on the sides/walls) other than the rounded ends in any typical furniture piece. Bear in mind that we are talking about a difference in the shape of the ends (with a slight void on the corners), not a huge gap between the mortise and the tenon. If a mortise is 2" wide/long and the tenon is 1-3/4" wide/long, of course, it is a structurally weak joint.

Some even make their own dominoes that are square (while the mortises cut by the joiner are curved): https://www.wwgoa.com/article/money-saving-diy-festool-domino-tenons/

Having said that, I would not intentionally leave a gap of 1/32" between the mortise and tenon, let alone 1/16", even for alignment purposes. Any misalignment (difference on the surfaces) due to a snug fit joint can be easily planed away.


Simon

[John TenEyck]

The glued side walls are what hold the joint together. Open space at the ends is irrelevant. If the glue fails it doesn't matter whether the tenon fits tight or loose, the joint will come apart unless it was pinned or wedged.


20+ years, no failures, including chair joints. Works for me.

John

[Brian Holcombe]

Maybe the glue will fail becuase the joint is being worked constantly rather than having an end grain abutment to fixture it.

[Brian Holcombe]

The strongest form of mortise and tenon joint, as far as I’m aware, is one in which the end grain abutments of the mortise are hourglass shaped. The tenon compresses as it passes through the joint and it must be literally separated from the continuing member to fail.

If we reduce this to two bits of side grain glued cross grain to two more bits of side grain the connection is initially strong but will certainly fail over enough time. Maintaining straight abutments does certainly give the joint something to counter the leverage being applied to it which will reduce the strain placed on cross grain sections.

When building a workbench I prefer a mechanically sound method which is not solely reliant upon glue. I use a workbench hours a day and have broken every weak point on it only to fix it with methods better prepared for what a workbench lives through.

[Jim Andrew]

I used the Schwartz book when I built my bench, and laminated the posts. Just left out a piece when laminating the posts to create the mortises. The mortises going at 90 degrees, I had to router out. I drilled holes and ran through the tenons, and glued them in so no way any joints can give.

[ChrisA Edwards]

I just did a few for my Plantation Shutters. I was going to use either my CBR7 kit or a horizontal router to do the mortises, but I opted for the Powermatic Mortiser. Expensive tool for one big project, but it was so easy to use, I'll use it for many future projects.

http://i1094.photobucket.com/albums/i452/cedwards874/Woodworking/Shutters/DSC_0242_zpshbfrwg9e.jpg

http://i1094.photobucket.com/albums/i452/cedwards874/Woodworking/Shutters/DSC_0247_zpswsyxebxl.jpg

[Edwin Santos]

I wonder if some of this hair splitting is theoretical. In a well fitted mortise and tenon joint, the side grain glue connection is usually more than strong "enough" (there's that word). So in direct answer to the OP, I'd agree with the instructor.

There's another factor to consider in this analysis. When I have made a perfectly mated blind mortise and tenon, with square abutments, I've had problems with a vacuum lock situation occurring during glue up that traps air in the bottom of the mortise and then makes the joint impossible to close because the air cannot evacuate. Due to this, I started leaving a tiny bit of breathing space at the top and bottom like John describes, to prevent this issue. With loose tenons you can cut a very shallow kerf down the length of the loose tenon stock to do the same.

When seeking ultimate mechanical strength like Brian is describing, if the design allows for it, drawboring the joint will make it indestructible IMO.
If it's a through tenon, wedging is another strategy. In fact, if you're really fastidious, you can slightly flare the through mortise and when wedged the tenon will basically create a large wedged dovetail. At that point, using glue is an option but not a requirement.

Edwin

[Simon MacGowen]

There's another factor to consider in this analysis. When I have made a perfectly mated blind mortise and tenon, with square abutments, I've had problems with a vacuum lock situation occurring during glue up that traps air in the bottom of the mortise and then makes the joint impossible to close because the air cannot evacuate. Due to this, I started leaving a tiny bit of breathing space at the top and bottom like John describes, to prevent this issue. With loose tenons you can cut a very shallow kerf down the length of the loose tenon stock to do the same.

Edwin

If I read John's post correctly, he was having the loose fit for alignment, not for the purpose you state here.

The proper way to avoid the "vacuum lock" you described is not to intentionally cut a tenon narrower but to cut the mortise deeper or the tenon shorter for glue escape, which will not compromise the structural integrity of an M&T. In fact, anyone who has worked with the dominoes knows that the mortises are always cut longer (deeper) than the dominoes for a similar reason. Dominoes are strong enough for hardwood doors as a loose M&T joinery.

For all intents and purposes, furniture joints will hold whether the mortises are round or square regardless of the ends of the tenons as long as the walls and tenon cheeks are a good fit and the gaps at the ends, if any, are not unreasonable for the purposes. Builds like benches or cabinets that are subject to racking and undue stresses, of course, require better care and attention.

Simon

[Mark Hennebury]

Tom,
This is a question not about whether the joint is strong enough, or whether it is necessary etc. that's irrelevant.
Nothing is necessary, you exist for a brief moment in time, make it count.
This is a question about who you are.
Do you know?
There are only two ways to do anything; the best, or the rest...... pick one.

[johnny means]

I wonder if some of this hair splitting is theoretical. In a well fitted mortise and tenon joint, the side grain glue connection is usually more than strong "enough" (there's that word).

Edwin

I've lap jointed 4x4s to build benches that would survive a lifetime of use in any shop. Modern adhesives and clean milling make all these arguments purely academic.

[Edwin Santos]

If I read John's post correctly, he was having the loose fit for alignment, not for the purpose you state here.

The proper way to avoid the "vacuum lock" you described is not to intentionally cut a tenon narrower but to cut the mortise deeper or the tenon shorter for glue escape, which will not compromise the structural integrity of an M&T. In fact, anyone who has worked with the dominoes knows that the mortises are always cut longer (deeper) than the dominoes for a similar reason. Dominoes are strong enough for hardwood doors as a loose M&T joinery.

For all intents and purposes, furniture joints will hold whether the mortises are round or square regardless of the ends of the tenons as long as the walls and tenon cheeks are a good fit and the gaps at the ends, if any, are not unreasonable for the purposes. Builds like benches or cabinets that are subject to racking and undue stresses, of course, require better care and attention.

Simon

I believe John and I are talking about the same process but two different benefits. The loose fit I mention is not in the cheeks of the tenon, but in the top and bottom edge, whether square or round. And it would be very slight, maybe 1/32 or less. I think John is talking about the space being for alignment, and I was suggesting a secondary benefit is to allow air to evacuate in the assembly process. I have noticed that the dominos have a crosshatch pattern in them which will also allow air to evacuate during assembly after which the compressed domino probably swells. I recall the domino machine will allow a setting for a slight bit of room for all these reasons. Although I agree that cutting a mortise slightly deeper or a tenon slightly shorter to allow a space for excess glue is a sound practice, but this space would not be nearly big enough for the air plunge from a super tight glue covered tenon.

Again, a lot of this is abstract. I think furniture design has more to do with joint failure than the joint construction minutiae we're discussing.

[John Goodin]

You mentioned you have limited experience making mortise and tenon joints and this is a great project to gain some experience and hone the skill. I, too and real actively inexperienced with the joint but find chisel work to be a nice diversion from the sound and sawdust generated by power tools.

[Brian Holcombe]

I've lap jointed 4x4s to build benches that would survive a lifetime of use in any shop. Modern adhesives and clean milling make all these arguments purely academic.

In what sense is it academic?

[Al Launier]

My experience with mortise & tenon joinery is very limited, so I really don't understand the round cornered mortise mating to a square cornered tenon as alluded to above. If the sides and ends of each are properly sized for a glue-up then how does a square corner fit into a round corner? I just see an interference fit that prevents assembly of the mating parts. I can accept a rounded tenon fitting into a square cornered mortise, but not the other way around. Perhaps my age is preventing me from understanding, but what did I miss from the above posts?

Also, I believe a round corner is less apt to fracture/split that a stressed square corner, so I've used the rounded corner for both the mortise and tenon.

[Mark Hennebury]

My experience with mortise & tenon joinery is very limited, so I really don't understand the round cornered mortise mating to a square cornered tenon as alluded to above. If the sides and ends of each are properly sized for a glue-up then how does a square corner fit into a round corner? I just see an interference fit that prevents assembly of the mating parts. I can accept a rounded tenon fitting into a square cornered mortise, but not the other way around. Perhaps my age is preventing me from understanding, but what did I miss from the above posts?

Also, I believe a round corner is less apt to fracture/split that a stressed square corner, so I've used the rounded corner for both the mortise and tenon.

This is how a square corner fits into a round and visa-versa


394409394410

[Simon MacGowen]

There are several stress tests done out there on joinery, with inconsistent results. Here is one that says the dominoes are stronger than the M&T: https://www.youtube.com/watch?v=xfjg3eQxlnw&feature=youtu.be&t=442

According to this test, it seems the cheek/wall fit is the critical factor. A domino joint does not have a perfect end-to-end fit between the domino and the mortise in the same sense of an M&T.

Simon

[Robert Engel]

Remember tenons are cross grain glue ups. I think the reason you see so many old ones pinned is the joints were more subject to failure.

So I agree with Pat people have success with various methods largely due to modern glues, not that its particularly a better way.

IMO most of the structural strength (not talking glue here) comes from nice beefy shoulders and a properly fitted joint.

[Simon MacGowen]

I have noticed that the dominos have a crosshatch pattern in them which will also allow air to evacuate during assembly after which the compressed domino probably swells. I recall the domino machine will allow a setting for a slight bit of room for all these reasons. Although I agree that cutting a mortise slightly deeper or a tenon slightly shorter to allow a space for excess glue is a sound practice, but this space would not be nearly big enough for the air plunge from a super tight glue covered tenon.

Again, a lot of this is abstract. I think furniture design has more to do with joint failure than the joint construction minutiae we're discussing.

For the record (official Festool manuals), the grooves on the faces of the dominoes are designed as glue pockets:
"The special shape of the DOMINO in combination with expanding glue pockets and lateral longitudinal grooves gives the dowels a secure grip."

In addition, anyone can check their Festool manuals and see that the narrow setting is recommended for tight joints; the wider settings are for alignments, not with glue escape in mind (because the dominoes are always shorter than the mortises). Anyone who wants max. strength (vs enough strength) should not use the wider settings, which are there for a different reason. The wider settings are perfectly fine if you are doing edge to edge glue-up or small boxes. Use them in a chair or bench, you have got a time bomb there.

I respectfully disagree with the thinking that "furniture design has more to do with joint failure than the joint construction minutiae..." A lot of joint failures can be attributed to poor execution, in addition to poor joinery choices (say, using a biscuit as corner joint for a chair). The joint construction itself does matter. For example, a loose fit of an M&T on the cheeks will remain weak even if the gap is glue-filled and the joint is pinned with dowels. All it needs to fail is the last straw.

Simon

[Jim Dwight]

My highest stressed mortise and tenon joints are in chairs based upon a design from WoodSmith. They have ladder backs held with mortise and tenons (5) and cross rails with mortise and tenons. All the tenons are 1/4 inch thick and the parts are fairly small. I've made lots of other mortise and tenon joints but the others are bigger if they take the kind of load that these get when a bigger person sits in the chair, or, even worse, leans back. There have been no issues.

The mortises were cut with a router and a spiral up cut bit. I use a template and have made 8 of these so far. I will probably make 6 more next year. All the mortises are rounded at the end and all the tenons are rounded to match. I do not deliberately cut them with extra clearance but occasionally have to trim a joint creating a little clearance. They all work so there must be a little strength margin despite the modest size of the joints.

I also have a hollow chisel mortiser and make square ended mortises with it, of course. I prefer them. But it is not practical to use them on these chairs. It is much quicker to tack the template in place and cut 6 mortises on each back leg in one setup. I could chop them square but I see no reason when it is much quicker just to round the tenons.

I am not trying to criticize anybody else's way of doing this. Just saying what I do with some success.

[Simon MacGowen]

My highest stressed mortise and tenon joints are in chairs based upon a design from WoodSmith. They have ladder backs held with mortise and tenons (5) and cross rails with mortise and tenons. All the tenons are 1/4 inch thick and the parts are fairly small. I've made lots of other mortise and tenon joints but the others are bigger if they take the kind of load that these get when a bigger person sits in the chair, or, even worse, leans back. There have been no issues.

The mortises were cut with a router and a spiral up cut bit. I use a template and have made 8 of these so far. I will probably make 6 more next year. All the mortises are rounded at the end and all the tenons are rounded to match. I do not deliberately cut them with extra clearance but occasionally have to trim a joint creating a little clearance. They all work so there must be a little strength margin despite the modest size of the joints.

I also have a hollow chisel mortiser and make square ended mortises with it, of course. I prefer them. But it is not practical to use them on these chairs. It is much quicker to tack the template in place and cut 6 mortises on each back leg in one setup. I could chop them square but I see no reason when it is much quicker just to round the tenons.

I am not trying to criticize anybody else's way of doing this. Just saying what I do with some success.

All the emperical info, therefore, does point to support what the instructor says (as quoted in the OP's post) - a rounded mortise and a square tenon is fine.

Simon

[John TenEyck]

All the emperical info, therefore, does point to support what the instructor says (as quoted in the OP's post) - a rounded mortise and a square tenon is fine.

Simon


Not surprising to me. Most of the test results I've seen always show that breaking strength is directly proportional to glued sidewall surface area. It will likely shock some readers that a bridle joint has higher breaking strength than a traditional M&T where the members are of the same size. The reason is simple; the bridle joint has a lot more glued sidewall area. Where the traditional M&T has a distinct advantage is that it lends itself to applications where no glue is used, such as in doors with wedged through tenons, or workbenches with tapered wedge through tenons that can easily be disassembled and moved. Also, glued M&T joints don't catastrophically fail if the glue fails, as a bridle joint would. But as long as the glue joint holds it makes no difference whether a tenon, round ended or square, fills the mortise completely or not. The sidewall surface area carries the load.

John

[Art Mann]

If I want to make all my moments count, I will not waste them practicing time consuming techniques that have no discernible effect on the end result. Your comment about "best" assumes that there is a common and well understood definition of the word. My definition of the "best" technique might be the one that meets all the aesthetic and structural requirements for a particular project with the least amount of effort. There are, in fact, many, many ways to create the joint under discussion and someone might find a reason to consider any of them the best.


Tom,
This is a question not about whether the joint is strong enough, or whether it is necessary etc. that's irrelevant.
Nothing is necessary, you exist for a brief moment in time, make it count.
This is a question about who you are.
Do you know?
There are only two ways to do anything; the best, or the rest...... pick one.

[Mark Hennebury]

If I want to make all my moments count, I will not waste them practicing time consuming techniques that have no discernible effect on the end result. Your comment about "best" assumes that there is a common and well understood definition of the word. My definition of the "best" technique might be the one that meets all the aesthetic and structural requirements for a particular project with the least amount of effort. There are, in fact, many, many ways to create the joint under discussion and someone might find a reason to consider any of them the best.

I agree with you that there are many versions and opinions of what the best is. I never suggested one for the OP or anyone else. Never stated that there is a common and well understood definition. From a structural standpoint there are many options that will do. My point was simple; the best or good enough, its about choice. Its about you, your personality and your reasons for doing what you do. You can do whats adequate for the job, if you do it for a living there is little reason to do more than is required. If you are doing woodworking because its a passion you can take it as far as you want. Many examples of incredible joinerywork around the world, developed over thousands of years by millions of dedicated people passionate about doing the best they can, "hardly necessary", and in my opinion hardly a waste of time. You waste your time whatever way you want, and i will waste mine my way. I happen to like joinerywork, I think that it is of value, and i will share my opinion with people anytime i can and i will encourage people to try to do the best designed, most precisely fitted joints that they can, but that is just me. What the best is, is for those who are interested to determine for themselves. The OP was concerned about the joint from the instructor, seemed to make him a little uncomfortable, which is promising to me. You can tell him its fine, good enough, i am going to tell him, he could do better.

[Brian Holcombe]

OP,

It’s worth a mention that in many workbench builds one can avoiding gluing the joinery at all and that is quite handy if you ever want to move the bench.

To secure the joint you can instead use a draw-boring method. Draw-boring can be knocked apart.

[Edwin Santos]

OP,

Draw-boring can be knocked apart.

Hi Brian,
I did not know this. I discovered draw boring only recently and I think it's awesome. The method I learned involved an offset hole (in the tenon vs. the mortise wall). By design, the offset pulls the tenon in very tightly. After the pins were driven in, they "deformed" in the course of negotiating the offset channel and I can see no way possible to drive them back out unless you can teach me what I'm missing.
Maybe in order to disassemble, the hole would not be offset? Thanks for any advice.
Edwin

[Larry Edgerton]

If you make through mortises they’re easier and stronger because you can cut from both sides toward center.

I like these and I like to taper the outside and drive in a wedge in a cut in the tenon.

[Brian Holcombe]

Hi Brian,
I did not know this. I discovered draw boring only recently and I think it's awesome. The method I learned involved an offset hole (in the tenon vs. the mortise wall). By design, the offset pulls the tenon in very tightly. After the pins were driven in, they "deformed" in the course of negotiating the offset channel and I can see no way possible to drive them back out unless you can teach me what I'm missing.
Maybe in order to disassemble, the hole would not be offset? Thanks for any advice.
Edwin

They deform on the way in and they deform on the way out. I’ve knocked apart 6” thick stock with 1/2” pins.

[John TenEyck]

A wedged through tenon like this makes more sense for a bench that needs to be disassembled.

https://lh3.googleusercontent.com/BhETmkye9DZp3TnlIHLXPACHxL09EN0LuxwTd5SEnio91iqFX_ 7PeFsQ_m6pEMIwUUxdLpLzXB-CnGJykctnM2wfiXvqKBV_4fjP6gzhdLfT_zxjSbB25xt3J-pjBBkZMztiz4rMiRYxrcrZ4dSAlbeErJR7lAo6M0xMHYC-qIfoPoaDyggVQuOuJ0Sesry-DKlm1QxxVKbm-_cwv1QUlBpEnr0zod7wNi8Co-6yeIt623s8W6KF9Vv2PuvoRvPXuFe6K1KTBucZeK94__sNZHXI HxzQQ6ruI1cVB-B3xIe1Uj4kwniFdMuM5ydByBHcwbEmFAk4ndMd1r_GYQxPd9Wo rO5S7gV1fqLvCbRSQPCxmOXmo2gnmhPlwkSPbuSdq3uosSNfUY PKORUDIjHGFeY9GtfpKl5gdTMuXlmXMPVlNYzjEVA6FLVWbsL9-9jXRWaCK07mBD_rlKuhSiKFFhKLHCpnSg4QgztS3kHpGv_afmo OuycgNDBNb1DoirdUNkmTYWrj-S72e4nuinwqVhOShGQSCFL9zSJ2CyLcbv-Q8tTKoIO1qlJsos7JUUzwO3-i58gJHnfKzxQmYvE5q0sOy444rqm_J-9L0y7gD_-NkKauYfAQOrK5arP9YaDCBDUjKNO_Xleahv7bBMPtlJFGqXjUl YeazunYt7Pd8bffVPOSGhgm6dXp1A=w1113-h626-no

I took this photo on a recent trip to Germany. It was in a wonderful town museum where my wife grew up. Her grandparent's apartment was now part of the museum! Small world. Anyway, you can see the wedged stretchers on the left side. The right hand side was held in a different, less obvious way. Why, I have no clue.

John

[Tom Porter]

Wow. That was a lot of information. Thanks for all the opinions. Before I make the bench, I'm going to experiment a bit trying different techniques to find what works for me. I'll incorporate much of what I learned here as I do that.

[Al Launier]

This is how a square corner fits into a round and visa-versa


394409394410

Thanks Mark, appreciate the time you took to display this. I'm glad you did as I would not have considered this concept as I feel all four sides should be in contact, especially since it wouldn't take that much more effort to make it this way. Although the shoulders are the prime means of preventing any rocking I feel the added contact with the tenon adds appreciably more.

[Edwin Santos]

For the record (official Festool manuals), the grooves on the faces of the dominoes are designed as glue pockets:
"The special shape of the DOMINO in combination with expanding glue pockets and lateral longitudinal grooves gives the dowels a secure grip."

In addition, anyone can check their Festool manuals and see that the narrow setting is recommended for tight joints; the wider settings are for alignments, not with glue escape in mind (because the dominoes are always shorter than the mortises). Anyone who wants max. strength (vs enough strength) should not use the wider settings, which are there for a different reason. The wider settings are perfectly fine if you are doing edge to edge glue-up or small boxes. Use them in a chair or bench, you have got a time bomb there.

I respectfully disagree with the thinking that "furniture design has more to do with joint failure than the joint construction minutiae..." A lot of joint failures can be attributed to poor execution, in addition to poor joinery choices (say, using a biscuit as corner joint for a chair). The joint construction itself does matter. For example, a loose fit of an M&T on the cheeks will remain weak even if the gap is glue-filled and the joint is pinned with dowels. All it needs to fail is the last straw.

Simon

Simon, I can see we disagree on some things. But I'll bet we agree on the following:

Sloppy, poorly fitted joints are never a good way to go.
The definition and criteria for "best" in joinery is personal, and perhaps circumstantial.

Edwin