Norm: I am ashamed of you

[Carl Eyman]

Today Norm was making his upolstered Foot stool. The legs were attached to the rails by dowels! Now I remember when I was making kitchen cabinet face frames in the 1950's and discovered dowels, I'd thought I'd found the ultimate joint. Since then I've learned that there is little long grain to long grain contact in such a joint. I've also seen many such a joint fail. So I was shocked when he remarked as he was assembling the dowel joint "That is a Stroonng joint." Granted the factory that was making the stools gave him the joint, but that's why so many pieces of factory furniture fail. Mea culpa, Norm. (PS. I am a big Norm fan, but wish he hadn't pulled this boo boo MHO)

[Jim Becker]

The upholstered Cigar Chair used double-dowel construction techniques.

[John Weber]

I think Norm used dowels because that is what the factory used. I watched the show and the joints looked plenty strong for the application. Plus a little change of pace isn't a bad thing either. If done right dowels work great in certain applications. I guess he could have used M&T, loose tenons, biscuits, pocket screws, sliding dovetails, or most any other jointery. He's used them all in the past. Looked fine to me.

John

[Carl Eyman]

The upholstered Cigar Chair used double-dowel construction techniques.Jim: I'd like to send you some teak dining chairs I've had in the attic for years because dowel and double dowel joints failed. Now, granted these short legs with generous butt joints even though end grain may last, but I still think with all the access he has to tools he could have used these new loose tenon things that look like several dowels side-by-side or some other alternative.

[Jim Becker]

No disagreement from me, Carl. I was only pointing out that the stool isn't the first project in recent years that Norm has used dowel joinery.

[William Lai]

Is there such a thing as an independent, authorative study on the various mechanical fastener system, such as dowels, pocket screws, biscuit? It's pretty confusing to me as a beginner to hear so many different opinions.

For example, these guys had an independent test that shows dowels are as strong as M&T: http://www.dowelmax.com/test_results.htm.

The other thought is that if you put enough dowels together on a joint, doesn't it approxmiate a M&T joint, or at least a floating M&T anyway?

[John Lucas]

Jim: "... these new loose tenon things that look like several dowels side-by-side or some other alternative."

I think you are referring to the beadLOCK system. They are loose tenons and work well. I am not a great proponent of dowels but they have there place and can work very well.

[Geoff Irvine]

There is an interesting chart on page 53 of 'Practical Design Solutions and Strategies' from 'Fine Woodworking' (ISBN 1-56158-344-8) about joining door rail and stile which is interesting. The author , John D. Wagner, indicates that in this situation joint 'strength' from strongest to weakest is :- 3biscuits, 2 biscuits, mortise and tenon = loose tenon, dowel (single) , tongue and groove then lag bolt. There doesn't seem to be much difference between the first four. This is not a rigourous scientific study but is an interesting reference article. Hope this helps.

[Michael Stafford]

The other thing that I noticed was that the wood used for the stretchers in the footstool was ambrosia maple...

I was surprised that he used dowels but maybe in that application it will not get the same wear and tear and stress that it would receive in a chair.

[Bob Smalser]

Is there such a thing as an independent, authorative study on the various mechanical fastener system, such as dowels, pocket screws, biscuit? It's pretty confusing to me as a beginner to hear so many different opinions.

For example, these guys had an independent test that shows dowels are as strong as M&T: http://www.dowelmax.com/test_results.htm.

The other thought is that if you put enough dowels together on a joint, doesn't it approxmiate a M&T joint, or at least a floating M&T anyway?

Sure...as long as the glue joint lasts. Problem is, it is bound to fail for two reasons:

1) Round tenons shrink seasonally to an oval, leaving almost zero contact surface, and in freestanding furniture the joint breaks under stress.

2) End grain is not a gluing surface at all and applying glue there at all is a complete waste of time. Crossgrain glue joints require face grain to be successful for more than a few seasons.

Dunno whether Norm doesn't understand any of that or he his so locked into a short-term commercial finish carpenter mindset, he simply isn't interested in teaching it, but "saving" much of his work 3 decades from now will be an expensive nightmare.

[Mark Singer]

Dowels are great in certain applications. In cabinet carcass construction they work very well, as used by Krenov, Frid, and Charlesworth (who uses them in the same way Krenov does). In frame joinery, they are not the best choice since the joints are subject to dynamic stress. By the time you use enouugh dowels, you really almost have a lose tenon. The stress on multiple dowels acts on the first and last dowel to a larger degree if they are in a series. So when those fail and are lose...the stress moves to the second and so on. The ones on the neutral axis only really provide shear resistance and not moment resistance which is the goal of frame joinery. For anyone who has made a Krenov style doweled cabinet can attest, aligning and transfering multiple dowels is difficult, even with a template and a lot of experience. For chairs joints should be Mortise and tenon with a vertical orientation. For strength Epoxy is the best glue in this application. Last year I returned several hundred chairs to a factory in Spain....the joints failed mostly from inferior glue and not enough glue. They used a finger jointed apron and finger jointed corner braces...which is a good joint, but without enough glue...it all faied and within a year, so the warranty was in effect. I did have to pay $5000 for shipping!

[Bob Smalser]

For chair joints ... Epoxy is the best glue in this application.

I wish everybody's Grandkids great luck in repairing these joints some day....because they are certainly gonna need it.

Epoxy is one of the few glues that can be made to work effectively (but not economically) on end grain (done in multiple coats allowed to cure in between)....and its flexibility and gapfilling are the cat's meow for fixed built-ins where dowel and other gizmo joinery are appropriate. Also great for round tenons, of course....except when they have to be dismantled.

For freestanding heirloom pieces, it's all just expensive preparation and storage of the next generation's fireplace kindling.

Build those pieces with traditional joinery and repairable glues so they can be repaired economically....because some day they are gonna need it.

It may not matter to you....but it probably will to a Granddaughter some day when somebody like me tells her it's a "time and materials" job because of the uncertainty of dismantling those epoxied joints or recutting all those failed Titebond joints.

[William Lai]

Thanks Mark on the explanation on the mechanical aspect of dowel construction due to stress. I can visualize the repeated cycle of weight on/off causing the dowels to graudally loosen on top and bottom, etc. In fact I'm sure I've seen it first hand.

One question though: Can you clarify what is meant by "Frame joinery", and is it only on chairs that you experience the above structural failure of dowels?

And just to finish off your thought: In cabinet construction, the forces are less due to the smaller weight involved, and the weight being more static in nature, and also less angular. What forces they experience are almost all shear forces, say in a hanging wall cabinet top and side joint. And in this case, all dowels are sharing the stress at the same time. Is that correct?

[Mark Singer]

William,

This is a very big subject and I will try to keep it simple. I will recommend to you and others, because these questions seem to always arise, Tage Frid Teaches Woodworking. It is a great resource and chosing the correct joint is one key to sucess as a woodworker.
Frame Joinery is where slender elements such as aprons, legs , stretchers are joined. This is different than carcass joinery, drawers or other elements. The slender elements are intended to transfer shear, bending, , tension and compression forces. In carcass joinery the elements are broad and there is a large surface to transfer forces. There is also "redundancy", meaning other elements within the cabinet help, if there is stress. The back for eample is a large shear element...it resists tortinal forces...so do other compnents such as partions, shelves...they work as a team.
In a table or chair the frame is subjected to dynamic forces where instead of redudancy , other elements can introduce forces which must be taken by a particular joint. Most frame joinery is based on some variation onf a M&T joint.

In a cabinet the sides with dowels or other joints and the back share the vertical loads...Here I am mostly concerned in stability so doors and drawers will fit well over time and through the seasons, without needing adjustment

[William Lai]

Thanks again Mark, for the explanation as well as the reference. I'll definitely add it to my reading list.