Dados vs Dominos (DF500)

[Packard]

"The test consisted of a single vertical panel with a horizontal shelf supported from one side only. They would then load the end of the short shelf to see when it would fail. No glue was used and they were only interested in what mechanical strength was added by the joinery method."

That a "hung" dado joint fared poorly in the lab test should be obvious to anyone even with little woodworking experience because in a typical 3/4" sheet good, the depth was only 1/4 to 3/8" or roughly one-third of the material's thickness. There is no way a friction joint is good enough for shelfing or cabinetry.

In the practical world, no one would use dadoes without glue (some even toenail it in addition to gluing it) in cabinetry work if strength is desired. As such, the Association's test is of no to limited value to me.

Yes. But RTA furniture rarely calls for glue. And some kitchen cabinets are site assembled. Sometimes, from my experience using hot glue, or minimal glue. So to the manufacturers, valuable information.

I always used glue, but it seemed like much of the side to side rackingstrength was derived from the backing board. With the other preferred joinery, the joints contributed to the structure. I’m sure dominoes would have fared as well as dowels in the tests.

 

[Packard]

About 8 to 10 years ago, IKEA, who funded much of the RTA testing, came out with an entirely new joinery system that does not rely on glue.

They call it a “wedge dowel” and I was able to find a video on that. Not for sheet goods, but an interesting peek into the corporate mindset in RTA manufacturing.

 

[luvmytoolz]

@Packard Very interesting video!

I do like how they think outside the square.

 

[Packard]

@Packard Very interesting video!

I do like how they think outside the square.

My concern with the wedge dowel would be that if it was assembled in the higher humidity summer, then the friction-fit would get looser come winter when the house gets drier.

And, of course a little bit of wood glue would go a long way. (I always glued-up my RTA furniture. Which worked against ease of packing and moving, but resulted in a more robust piece of furniture.

 

[luvmytoolz]

My concern with the wedge dowel would be that if it was assembled in the higher humidity summer, then the friction-fit would get looser come winter when the house gets drier.

And, of course a little bit of wood glue would go a long way. (I always glued-up my RTA furniture. Which worked against ease of packing and moving, but resulted in a more robust piece of furniture.

That definitely wouldn't be my method of choice for the reasons you mention and because I like to over-engineer things, but if I was to utilise it, I'd be putting ribs on the back-edge and the last thing would be to fit a removeable block in the recess behind the wedge to keep it firmly locked in place.

Still, the fact they had the guy spend a month finessing a design I thought was good on the part of corporate. My workplace it'd be "we want it yesterday or else!".

 

[Mini Me]

My concern with the wedge dowel would be that if it was assembled in the higher humidity summer, then the friction-fit would get looser come winter when the house gets drier.

And, of course a little bit of wood glue would go a long way. (I always glued-up my RTA furniture. Which worked against ease of packing and moving, but resulted in a more robust piece of furniture.

In my part of the world the humidity does not vary to a huge extent and would generally not be a concern. My view on all this is that hobbyists using commercial techniqu.es don't get the rewards and satisfaction that traditional wood working gives you. I was involved in a traditional woodworking school not as a teacher but other matters along with being a pupil and it soon became obvious that most pupils wanted to learn everything in a few lessons so I can see why machines like the Domino appeal as they shortcut the journey into woodworking. The Domino like a lot of machines saved a lot of time for the smaller companies doing semi mass production and we hobbyists took it up to avoid doing traditional joint designs which require a longer time to learn.

 

[ChuckS]

"My view on all this is that hobbyists using commercial techniqu.es don't get the rewards and satisfaction that traditional wood working gives you."

I draw as much satisfaction (reward?) from using power tools, including the DF500, as from using my hand tools to complete a project. I have done many many joints with a handsaw and chisels, but I feel no less happy when something is made with floating tenons or finger joints. I'm sure I'm not speaking just for myself.

 

[luvmytoolz]

"My view on all this is that hobbyists using commercial techniqu.es don't get the rewards and satisfaction that traditional wood working gives you."

I draw as much satisfaction (reward?) from using power tools, including the DF500, as from using my hand tools to complete a project. I have done many many joints with a handsaw and chisels, but I feel no less happy when something is made with floating tenons or finger joints. I'm sure I'm not speaking just for myself.

The more tools the better in my book! ;-)

 

[mino]

...
I forget which one said, “regardless of the type of fastener, no fastener should be within 2” of the end of the board. So a 12” wide board would have 8” of available space for fasteners. This because the board needs space on both sides of each fastener to prevent the board from blowing out.

To me, the tests made more sense for the RTA industry than the kitchen cabinet industry. Most cabinets are fastened to the wall and the wall becomes part of the structure greatly improving resistance to racking. Ready to Assemble Furniture is generally free-standing and racking strength would be essential.
...

There are two base approaches to a structure design:

A) Component + Joints /a gravity-aligned compound assembly basically/
- generally the joint is the purposely weak point, this means there are major stresses at the joint location, necessitating compensation by leveraging the elasticity of the component /like the 2" mentioned/

B) Single-structure /a single structural element, even if originally made from components, gravity-independent mostly/
- here the joints are stronger or comparable to the strength of the material they attach to, this means there are no major stresses concentrated at the joints, the material stressing is spread around the structure and is defined by geometry, not joint locations.

When you glue stuff with /many/ dowels, you create a single-structure piece. At that point it is irrelevant if the dowels /dominoes/ are 1/4" from the edge or 1". To the contrary, having them 2" out is undesirable as those two inches of shelf are then "unsupported" which you definitely want to avoid in a single-structure piece as it becomes the weak point.


Long story short, one cannot really transfer RTA techniques which are universally of the compound assembly type to single-structure designs and vice versa. Some techniques technologies are applicable to both, but e.g. a free-floating dowel used in RTA is not to be confused with a glued dowel. Two completely different joints.

---
ADD:
One thing to add. Those "tests" published by the various "associations" are PURPOSELY made to mislead the audience. Their purpose is NOT to test anything but to convince /uninformed/ public on "how good they are making the furniture". E.g. that dry dado "joint" was included in the "test" only to create FUD and make the public think a custom-furniture using dadoes is inferior to "proper" industrial made one. Its inclusion is a dead giveaway it is no test but a false advert made up by marketing specialists. No I am not joking. That is the /false/ message a non-engineer takes from such "tests".

These test get commissioned after some study finds why people mistrust your product. Based on that there is a contract given out to make an "education material" to "educate" the customers in the direction you want. Such a "test" needs to be reasonably trustworthy, so often proper scientific methods are used so even the more-technical but non-involved customers cannot "attack" the test itself. The lie is not in the method to test, it is in what is being tested. Imagine if a non-tech wife is presented with such a "test" along a woodworking husband. The husband will be unable to convince her the test is bull on merit, even if he immediately notices that(!). She may defer to him, but would still think "the scientists" are right.

Basically, a PsyOp. Given even some folks here fell for it, discussing it seriously, a pretty effective one at that.

 

[Packard]

There are two base approaches to a structure design:

A) Component + Joints /a gravity-aligned compound assembly basically/
- generally the joint is the purposely weak point, this means there are major stresses at the joint location, necessitating compensation by leveraging the elasticity of the component /like the 2" mentioned/

B) Single-structure /a single structural element, even if originally made from components, gravity-independent mostly/
- here the joints are stronger or comparable to the strength of than the material they attach to, this means there are no major stresses concentrated at the joints, the material stressing is spread around the structure and is defined by geometry, not joint locations.

When you glue stuff with /many/ dowels, you create a single-structure piece. At that point it is irrelevant if the dowels /dominoes/ are 1/4" from the edge or 1". To the contrary, having them 2" out is undesirable as those two inches of shelf are then "unsupported" which you definitely want to avoid in a single-structure piece as it becomes the weak point.


Long story short, one cannot really transfer RTA techniques which are universally of the compound assembly type to single-structure designs and vice versa. Sime techniques are applicable to both, but e.g. a free-floating dowel used in RTA is not to be confused with a glued dowel. Two completely different joints.

Interesting. Unfortunately I have not been able to locate the two studies on line (though that is where I found them originally). So I don’t know which (RTA or Kitchen cabinet) study mentioned the 2” bit. My jigs space the holes at 32mm (1-1/4” approx.). I will check when I get home to see if the initial hole is 32mm from the edge. It might be slightly more than 32mm, but in my memory not nearly 2”.

I recall Nikon camera used tiny brass screws in their assembly, when stronger stainless steel screws were available at a similar cost. Of course they did that because if the camera was damaged the brass screws were cheaper to replace than the castings they were attached to. So they used the weaker screws to be sacrificed to protect the castings.

In any case, my design and engineering skills are of the that-looks-like-it-is-strong-enough school of design. I am almost certain that I over-build almost everything I make.

One notable exception: As an experiment, I built a medicine cabinet using “Internet construction”. That is butt joints, nails or screws and glue. With the medicine cabinet, that meant butt joints and 18 gage nail gun + glue. At the end, it was surprisingly sturdy. I have not gone back to that type of construction, but it did impress me.

What prompted the experiment was my demolition of a 1950s site built bathroom vanity. All butt joints, no glue, and hammer and nails. It was amazingly resistant to being dismantled.

 

[ChuckS]

"E.g. that dry dado "joint" was included in the "test" only to create FUD and make the public think a custom-furniture using dadoes is inferior to "proper" industrial made one."

I really have had a hard time finding or remembering seeing any woodworking vidoes or knowing any woodworkers where a dado joint is used without glue in a ply construction. If the dado is cut and used as a slot and not for structural strength, that's a different story.

Why would not any serious kitchen cabinetmakers challenge that dry dado test element when they saw it? Maybe some did.

 

[Packard]

"E.g. that dry dado "joint" was included in the "test" only to create FUD and make the public think a custom-furniture using dadoes is inferior to "proper" industrial made one."

I really have had a hard time finding or remembering seeing any woodworking vidoes or knowing any woodworkers where a dado joint is used without glue in a ply construction. If the dado is cut and used as a slot and not for structural strength, that's a different story.

Why would not any serious kitchen cabinetmakers challenge that dry dado test element when they saw it? Maybe some did.

I also cannot imagine using a dado without glue. But from the perspective of the RTA and kitchen cabinet manufacturers, this was probably useful information.

I do see (In Lowes) production cabinets with face frames attached to the boxes using dados. The dado is in the face frame and the box wall is inserted at full thickness. I think once you have a fixed setup for this type of joint, that it is probably quite efficiently done.

Lowes version used either 1/2” or 3/8” thick stock for the boxes, but the attachment is the same:

 

[ChuckS]

I also cannot imagine using a dado without glue. But from the perspective of the RTA and kitchen cabinet manufacturers, this was probably useful information.

I do see (In Lowes) production cabinets with face frames attached to the boxes using dados. The dado is in the face frame and the box wall is inserted at full thickness. I think once you have a fixed setup for this type of joint, that it is probably quite efficiently done.

Lowes version used either 1/2” or 3/8” thick stock for the boxes, but the attachment is the same:

View attachment 380351

One source to the image you produced is here: https://www.woodmagazine.com/woodwo...the-face-frames-first-for-my-kitchen-cabinets

The article doesn't say glue is used or not used with the dado bottom. If a box or cabinet is screwed to the wall, glued or unglued dado may not make much a difference. But the dado discussion we're having here is about free standing cabinets and the dadoes need to be glued to add strength (see OP's and 2-meter long shelving).

You said the association's hung dado test was useful to kitchen cabinetmakers. How? Did some kitchen cabinetmakers or installers use friction fit shelving (with one end unsupported) somehow?

 

[Packard]

One source to the image you produced is here: https://www.woodmagazine.com/woodwo...the-face-frames-first-for-my-kitchen-cabinets

The article doesn't say glue is used or not used with the dado bottom. If a box or cabinet is screwed to the wall, glued or unglued dado may not make much a difference. But the dado discussion we're having here is about free standing cabinets and the dadoes need to be glued to add strength (see OP's and 2-meter long shelving).

You said the association's hung dado test was useful to kitchen cabinetmakers. How? Did some kitchen cabinetmakers or installers use friction fit shelving (with one end unsupported) somehow?

Do I think dadoes are generally a satisfactory joint? Yes.

Do I believe that dowels, Confirmat screws or dominoes offer better side to side racking strength? Yes, primarily after considering the test results.

Dadoes offer excellent load carrying capability. If sufficient number of fasteners are used, they will too.

But dadoes, no matter how well executed, or we’ll glued will never match the racking structure that the fasteners supply.

Dado-built cabinets or furniture depend on the backing board or the wall for that structure. And those are rarely glued.

 

[mino]

Do I think dadoes are generally a satisfactory joint? Yes.

Do I believe that dowels, Confirmat screws or dominoes offer better side to side racking strength? Yes, primarily after considering the test results.

Dadoes offer excellent load carrying capability. If sufficient number of fasteners are used, they will too.

But dadoes, no matter how well executed, or we’ll glued will never match the racking structure that the fasteners supply.

Dado-built cabinets or furniture depend on the backing board or the wall for that structure. And those are rarely glued.

II will try some pointers that may give you food for thought:

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by a dado? Even glued, not mention unglued.

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by confirmats that was expected to carry any load to speak of, i.e. was not purely decorative?

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by unglued dowels + screws that was expected to carry any load to speak of, i.e. was not purely decorative?

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by screws that was expected to carry any load to speak of, i.e. was not purely decorative?


None? Well, there is a reason. Only an idiot would design such over-extended piece as it is pretty much guaranteed to fail even if it could hold itself together to not fall-apart immediately.

Now why do you think someone is 'testing' such a "joint"? Do you honestly believe they are all idiots in that association?

Hint:
They are 'testing' such a joint scenario because it is one of the few scenarios where dry dowels + screws and confirmats can "win" on relative*) strength against other joinery options. Who cares no sane cabinetmaker, including the RTA members, would ever use such a joint.

*) indeed, they are the strongest within the use case, but still fail the use case, making the test completely irrelevant while most customers would be unaware of such "detail".

 

[mino]

On the OP post, I think enough was written by folks more experienced than me on the practicality.

I will comment on the strength part:

As it was mentioned, as far as load-bearing strength goes:

A dado, when done well, will be inherently superior to ANY other joint as the whole end of shelfs is supported. This holds true as long as the shelf is not overloaded - if overloaded/under-specced, a shelf that bends too much can be "pulled out" of a dado. This can be a problem with chipboard or MDF as those are "liquid" to a certain extent and sag/bend gradually over time.

This is why dadoes should be used with caution with junk materials like MDF/chipboard shelving or backed up by screws for safety. Once ply or, especially, wood is involved, the advantage moves to dadoes as unlike with dowels/dominoes there is no risk of chipping the shelf when overloaded and too-few dowels/dominoes.


But there are other aspects to strength - the mentioned "rocking strength" etc. When one compares to dominoes/or dowels/or confirmats each of these is vastly weaker on load bearing BUT that is generally a non-issue as the limiting factor is usually the shelving board, not the joint itself. /2 meter shelving excepted .../

On the rocking strength a glue-only domino carcass is weaker than a dado-ed one BUT it has a bigger hysteresis at the same time. Meaning, a domino-jointed carcass will start failing sooner, but it will fail gradually and it will be more resilient to humidity-induced failures. All things considered, it is a wash. More depends on care/technique than the joinery type used.


End of the lesson for me, when I was exercising my mind on these things:
Dadoes are great for face-frame designs with actual wood and ply as they avoid the 'unclean edges' problem. Dowels/dominoes are great for frameless designs. I would not consider either superior/inferior in general. There are use cases for different joinery options and the more in one's disposal, the more optimal one can choose. As for efficiency, CRG made all the key points so will not repeat.

 

[luvmytoolz]

For any type of cabinet or shelving I always use dado's extensively and glue everything. The only time I don't glue material in a dado is when it's a slot (usually dovetailed) to allow expansion for a solid timber panel (usually a cabinet top).

For items that will get a full sized raised panel rail/stile fitted to the sides, I also screw the shelves in through the dado before the panels are fitted.

 

[ChuckS]

Do I think dadoes are generally a satisfactory joint? Yes.

Do I believe that dowels, Confirmat screws or dominoes offer better side to side racking strength? Yes, primarily after considering the test results.

Dadoes offer excellent load carrying capability. If sufficient number of fasteners are used, they will too.

But dadoes, no matter how well executed, or we’ll glued will never match the racking structure that the fasteners supply.

Dado-built cabinets or furniture depend on the backing board or the wall for that structure. And those are rarely glued.

Sorry, I'm still not sure from your detailed reply how the friction-fit dado test was useful to a kitchen cabinetmaker. Perhaps it was not fair to ask that question to you as you didn't design or run that test. I'll leave it at that.

 

[Packard]

II will try some pointers that may give you food for thought:

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by a dado? Even glued, not mention unglued.

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by confirmats that was expected to carry any load to speak of, i.e. was not purely decorative?

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by unglued dowels + screws that was expected to carry any load to speak of, i.e. was not purely decorative?

How many pieces of furniture did you see with a single-sided shelf/board 'secured' by screws that was expected to carry any load to speak of, i.e. was not purely decorative?


None? Well, there is a reason. Only an idiot would design such over-extended piece as it is pretty much guaranteed to fail even if it could hold itself together to not fall-apart immediately.

Now why do you think someone is 'testing' such a "joint"? Do you honestly believe they are all idiots in that association?

Hint:
They are 'testing' such a joint scenario because it is one of the few scenarios where dry dowels + screws and confirmats can "win" on relative*) strength against other joinery options. Who cares no sane cabinetmaker, including the RTA members, would ever use such a joint.

*) indeed, they are the strongest within the use case, but still fail the use case, making the test completely irrelevant while most customers would be unaware of such "detail".

If you wanted to test for racking (side to side) strength of a joint, what test would you use? I think the single sided, cantilevered shelf was a very reasonable choice, not as a final way to assemble a product, but as a test of relative strength of the joint.

I would note that the RTA furniture would not use glue normally.

I would also note that some of the cabinets sold to commercial installers are of the “flat pack” variety, that is they have to be site assembled. And if second-hand cabinets that my friend purchased (looked nice) are typical, all the joints were bonded using a couple of drops of hot glue. I know this because he asked me to repair the drawers. A close look showed that all the joints were hot glued. So the type of joint probably was as important as who assembled it.

My early book shelves used dadoes to support the shelves (glued), and a backing board slid into grooves at the rear of the sides. I attached the backing board using a few nails across the top shelf and bottom shelf. At the time, I was copying what I saw in purchased shelf units. In retrospect, a rabbet at the rear of the uprights and glue would have made a stronger piece of furniture. The backing board provided the bulk of the racking strength.

As RTA, dowels plus 1/4 turn fasteners, or confirmats would be sturdier than dadoes.

In summary (my opinion), dados excellent at two things. Guaranteeing that the left and right side supports are at the same height, and load carrying. They are deficient (compared to other methods) in racking strength. And I believe you can fracture the glue’s hold in a dado by rocking the joint side to side.

So, pick dadoes when you need substantial load carrying for the shelves. Pick another joint for racking strength.

I would note that commercial testing frequently relies on an “absurd” environment to test for relative performance.

Take, for instance, the test for corrosion (ASTM B117). They place a plated or painted object inside a heated chamber and mist saline solution into the chamber. They check for corrosion in 2 hour intervals.

That is not a realistic environment for any product anywhere. But what it does is give a relative, accelerated result. So you can compare pre-galvanized materials which comes in at about 40 to 45 hours with zinc plated materials which typically test at 110 to 120 hours or powder coating which ranges from 200 to 400 hours.

None of those results tell you how long a product will last in real life, but will tell you that zinc plated parts will last a lot longer than pre-galvanized parts, and powder coated will last far longer.

So the cantilever test does not represent a real life situation; it gives a relative test of strength for the various joints. The dadoes did poorest. The dowels did near the top and the Confirmats won the race, offering a combination of the strength of the dowels with the holding strength of a screw. Unfortunately it is a visible fastener and is often unacceptable.

In any case, it is a well-established fact that testing does not have to replicate real life environments to be valid. It only needs to test for the criteria that was being tested.

 

[ChuckS]

"So, pick dadoes when you need substantial load carrying for the shelves. Pick another joint for racking strength."

Someone using a dado joint as a core means to resist racking? That's news to me.