Thread: DIY Furniture: Dowelmax or Festool Domino?

Its posts like this that make the forum more interesting. We are all big and ugly enough to handle a bit of stick and healthy debate. Look at the viewing numbers compared to the other posts of the week Anyway the Domino is heaps better than anything else

Has anyone done any work today?

Originally Posted by Ron Dunn

No, I don't, which is why I was interested in Ian's response.

I think his opinion is that something made with a modern jig is a loose tenon, not a dowel. That's an interesting perspective, but I'm not sure where I'd draw the line between tenon and dowel.

The reading I've done - old books, magazines, etc. - tended to show minimum dowel recommendations. I've seen similar numbers to those quoted by Ian as minimums, but I can't recall ever seeing a maximum.

Ian, do you have any opinion on why the number of dowels should be limited?

Ron

I'm currently doing the Night trade Cabinetmaking Trade course at Lidcombe TAFE which uses what I'm told is the National Certificate IV curiculum.

we're taught that:

• in a commercial enterprise, calculating the number of dowels required is matter of economising on time and material.
  • more dowels take more time so cost more
• dowels are sized on the basis of ½ the material thickness — so 18mm board uses 9mm dowels, a 12mm board 6mm dowels, a 30mm thick rail 15mm dowels
• dowels are spaced on the rule of thumb of one dowel per inch of sawn width (dressed dimensions are always less)
• you need a minimum of two dowels to prevent twisting, so if you have to you would use a smaller dowel than suggested by the ½ the thickness rule (round chair rails can twist as much as they like so only "need" one dowel)
• you rarely need more than 3 dowels per joint — typically a 90mm table apron would get 3 dowels although the rule of thumb suggests 4
• dowels are cheaper because you use less material (a chair rail for example only needs to be as long as the space between the legs, not that space plus extra for the tenons both ends)
• dowels are quicker than M&Ts — in batch production you would always use a template for both faces of the joint, so drill 4 holes wipe in some glue push in the dowels assemble and clamp the joint and your done — there's no need for a trial fit

the other thing we're taught about dowels is that they are a form of loose M&T — loose M&Ts are basicaly defined as any butt joint with internal reinforcement, examples that readily come to mind are:
beadloc joints
Domino joints
dowel joints
DowelMax joints
"true" loose M&T where the mortices would typically be cut with a dedicated morticer or a router

biscuits, however are considered a form of spline joint
spline joints are also reinforced butt joints but unlike a loose tenon, a spline is recognised by its width being much greater than the depth of the slot cut in the matching pieces.

and pocket screws are a different class again


Now, I'm deliberately making a distinction between dowelled and DowelMax joints.
One of the posts above talks in terms of using 6 or 8 or 10 or 12 compressed dowels per DowelMax joint.
I'm not saying this is "wrong" but when you start to fill up a joint with that much bridging material it's a bit rich to call it a "dowel" joint


ian

Originally Posted by ian

Has anyone done any work today?

Originally Posted by Ron Dunn

the joint under examination is a large dowel (say 1" diameter) that joins a chair rail to a chair leg. I say it is extreme because of the joint stresses (ie, rocking on a chair), the size of the dowel, and the singular nature of the dowel.

It is really an excellent article. Even if you're not a subscriber, if you have the interest to look it up in a library or somewhere it was from Fine Woodworking #21.

Hoadley made a number of interesting recommendations based on his testing to create a better joint:

1. Make the mortice (he uses that term) depth at least 1.5 times its width.

2. Dry your stock to less than its final equilibrium before construction.

3. Drill carefully, creating a smooth mortice.

4. Orient the dowel grain in line with the stock grain where possible.

5. Finish the final product, don't leave it raw.

He further examines the use of glues, recommending further experimentation with silicone-based glues in joints under compression/expansion stress. He found that when he (a) split the tenon [dowel], and (b) used a silicone glue, he had no joint failures in his experimental set of school library chairs, compared with one joint failure using white-glue and the 5 steps described above.

I'm sure I've left out points that others would find interesting/important, but to me, the crucial thing is that Hoadley was not saying that dowelling sucks, but describing common problems and resolutions for the style of joint described.

Ron

thanks for posting that summary

interesting, that bit about drying to less than final equilibrium, this is what Windsor chair makers use
they dry their rails to less than equilibrium (traditionally by poking the tenon in a fire) and then stick them in a hole drilled in a green leg
the leg dries to equilibrium so the hole gets smaller in diameter
the rail absorbes moisture from the leg so gets bigger in diameter
resulting in a very tight joint that doesn't require gluing

so a joint made as a tight fit for a very dry (or compressed) dowel would be stronger regardless of which glue was used by virtue of the lock achieved when the dowel takes up moisture and expands


ian

Originally Posted by ian

loose M&Ts are basicaly defined as any butt joint with internal reinforcement, examples that readily come to mind are:
beadloc joints
Domino joints
dowel joints
DowelMax joints
"true" loose M&T where the mortices would typically be cut with a dedicated morticer or a router

At the risk of reinforcing the Dominonazi tag that I have been branded with what`s the difference between a “true” M&T and a Domino join?

It's simply "horses for courses", what suits you best is what you use.

Original poster had already made decision, they just wanted to promote a product.

Try telling a hand cut dovetail maker that using jigs are just as good to see similar results.

after flicking through this thread at the various points made, I have come to the conclusion that we are an endangered species.....the skills acquired in using a simple mallet and sharp chisel, pencil and set square, are being lost to the high tech gadgets of today .......

Tuesdays Time Team on ABC dug up a Roman river barge on the Rhine with Domino's used to hold it together. Dated to around AD 85 +/- 6 years

Sure it was done with hand tools but structural strength cannot tell if it was hand or machine cut.

Originally Posted by Lignum

At the risk of reinforcing the Dominonazi tag that I have been branded with what`s the difference between a “true” M&T and a Domino join?

Lig
maybe I should have said "traditional"

by "true" or "traditional" I mean a loose M&T where the tenon is neither compressed nor mass produced, but rather sawn from similar or stronger stock to that in the joining pieces (with the edges rounded or left square depending on the asethetic of the maker and the method of cutting the mortise)


yes I know you can make your own domis, beadloc, etc


ian

Originally Posted by BobL

Humm . . . I don't know where you're buying your HSS bits from but I just grabbed 3 HSS bits out of our workshop and went up to the Materials testing lab and measured the Rockwell C hardness.

Wow. It must be great to have those capabilities at hand Bob. Where do you work that you have a Rockwell tester?

Originally Posted by Ron Dunn

US-Oz, I'm not sure there is actually a greater surface area in a dowelled joint.

The surface area of a dowel is: Pi * diameter * length

For a maximum sized dowel of 10mm, inserted to the typical half-length of 25mm, each insertion has a surface area of around 785 square millimetres. Thus, five dowels would have a combined surface area of 3,925 square millimetres.

If a tenon was cut to occupy the same outer dimensions, it would have a surface area of (2 * cheek area) + (2 * edge area). The maximum comparable size of the tenon would be approximately 80*10mm * 25mm depth, so the surface area would be 5,600 square millimetres.

In practice, it would be easy for the tenon to be made longer, which wouldn't be easy or practical with store-bought dowels. Thus a larger glue surface again.

I hope I got the maths right

Sorry I missed this post, Ron.

The surface area of a cylinder like a 10mm x 50mm dowel = (area of top) + (area of bottom) + (area of sides)
= (2pi*radius-squared) + (2pi*radius*height)
= 1729sqmm
halve that for the one side of an insertion, so 864sqmm

For 5 dowels, that's 4320sqmm

Now you posit a tenon as wide as the 5 dowel settings, which is pretty wide, and it would indeed give a larger area for glue contact. But it certainly isn't what's occurring with the Domino, which would have a lesser contact area since it uses small tenons, usually one or two. So Dowelmax would have the gluing area advantage over the Domino.

But the one thing that is constantly thrown in the face of the "dowel camp" is the supposed grain to grain advantages of flat cheeks. I contend this is not nearly the huge advantage people are making it out to be, as is shown by the various tests that prove the multidowel joint beats the Domino joint hands down. Now why is this?

It's because modern polyurethane glues actually glue better with end grain available than without. End grain sucks up these glues to make a stronger bond. I refer you to the post I made above quoting the makers of Gorilla Glue, Excel Glue or Tuff Glue where they claim that the glue is "sucked up into the end grain wood tubules", unlike the more common aliphatic glues (PVA, White, Yellow glues) of old.

Now is there any scientific evidence that wood grain direction and contact are much less important than they used to be? Well yes, there is:

Originally Posted by Journal: Wood Science and Technology

Effect of grain angle on shear strength of glued end grain to flat grain joints of defect-free softwood timber

Jürgen Follrich1 Contact Information, Alfred Teischinger1, 2, Wolfgang Gindl2 and Ulrich Müller1
(1) Wood Kplus – Competence Center for Wood Composites and Wood Chemistry, Linz Austria, Peter Jordan Strasse 82, 1190 Vienna, Austria
(2) Department of Material Sciences and Process Engineering, Institute of Wood Science and Technology, BOKU – University of Natural Resources and Applied Life Sciences, Vienna, Austria

Received: 4 August 2006 Published online: 19 April 2007

Abstract

To evaluate the effect of grain orientation on the adhesive bond strength, three-layered Norway spruce wood specimens were tested in shear. The two axial-oriented outer layers were jointed with the middle layer using three typical glues for load bearing constructions, i.e. one-component polyurethane (PUR), melamine–urea–formaldehyde (MUF) and phenol–resorcinol–formaldehyde (PRF). The grain orientation of the middle layers was varied from 0° (parallel to grain of the surface layer) to 90° (perpendicular to grain of the surface layer) in incremental steps of 10°. Samples with middle layers oriented parallel to the outer layers showed shear values in the range of solid spruce wood. Decreased shear strength values were expected for increased grain angles of the middle layer. However, no explicit tendency was observed for the shear strength in dependence of the grain direction. In general, MUF-bonded samples showed slightly higher shear strength values compared with PUR- and PRF-bonded specimens.

I hope this shows definitively that the arguments that raged in the thread above are largely irrelevant and outdated. A multidowel joint is stronger than a Domino joint, of that I am sure, not only because the tests prove that to be true, but also because of the clear scientific evidence in that direction. People here are quoting ancient lore like Hoadley, not realising that a lot of what he said was based on vastly inferior glues (and tools).

This "end grain phenomenon" may also explain why some of these tests are showing, counter-intuitively, that the multidowel joint is stronger not only than the Domino, but also than the traditional M&T joint!

US-Oz, we've got the same maths, but I think your logic is a little out ... you can't take the end-surfaces into account.

One end will not have any glue at all, being at the barrier between the two surfaces to be joined.

The other end will be in the bottom of the drilled hole. It is highly unlikely that this end of the dowel will be in contact with a gluable surface, either because of the shape of the drill bit, or the depth of the hole being longer than the dowel.

The round surface area is about as good as it gets. And even then, we've not taken into account any effect - good or bad - of fluting in the dowels.

Like I said earlier ... when I find a spare $10million I'll fund some research in the field

You've had all day and that's the best you can come up with? Disappointing...

Mate, if you want to use polyurethane glues then go for your life. I'm over joyed for you. You will have the strongest furniture this side of the black stump. Oh that's right, you're on the other side of the black stump, aren't you? But, again, I ask you, if the glue is strong enough to join end grain to long grain, why do you need dowels? Why don't you just slap on some glue and clamp it together? Can't get quicker than that and if what you say is true, it should be just as strong. The dowels are superfluous if you believe your glue hyperbole.

So how long have you been using polyurethane glue with your furniture? Have you found any problems with handling it? And how are you going with sourcing your compressed dowels?