Exterior Door Construction Details

A discussion of the relative merits of mortise and tenon versus dowel construction. July 3, 2006

We have the heads to rail and stile exterior doors. I have never made them, but have repaired some. What method other than M&T would be acceptable? We have thought of dowelling them, but that might get a bit tricky lining them properly. Is there glue or epoxy that would give as good a joint with just the cope? Would like to be able to give a 10-15 year structural warranty.

Forum Responses
(Architectural Woodworking Forum)
From contributor B:
You could drill oversize holes and use threaded rod with epoxy. I've drilled 1/2" holes with 3/8 allthread and secured it with West System epoxy. Worked great and still holding after 12 years. The cope will line up the rails and stiles horizontally - watch the vertical, as there is some slop in the oversized holes.

From contributor N:
Traditional door construction, meaning m&t, is crucial to the long term integrity. First, the bottom rail tenon is the dynamic component in the door sagging. Next, the waist rail tenon has the same purpose. If you notice older large entrance doors, you will see very large bottom rails. This is both for aesthetics and the ability to cut a tailer tenon. Once I had the idea of doweling, and standing behind the product has always stopped me.

From contributor D:
Mortise and tenon is the traditional and, in my opinion, the best way to go. Doweling, properly executed, will work well with the cope and stick, but must be accurate and true. Do not rely on just the cope and stick and glue to hold a door together, much less an exterior door.

As to the structural warranty, it is always dependent upon the exposure an exterior door will receive, so blanket coverage is difficult. If the exposure is not a serious problem, then proper materials and construction will yield a door that will easily outlast you. A product life of 10-15 years implies disposable woodwork, therefore disposable resources, time and effort. The wood we use and the pieces of our lives we sell are too dear to be disposable. Build your best. If the customer can't/won't afford it, let them buy plastic.

From contributor R:
Would it be feasible to use mortise and loose tenon? I've been contemplating the same issues, and it seems that one could cut rails and stiles to length, mortise the stile and end of the rail, use cope and stick cutters to profile the whole thing, and then insert loose tenons on assembly. If so, that would certainly lend itself to my existing tooling.

From contributor J:
I guess I'm having a hard time understanding how the cope cut is made with a tenon sticking out? How many of you use M&T in stave core door construction?

From the original questioner:
I do not want to sound like a wood butcher, and I certainly want and need the doors to last longer than "disposable woodwork," but they must be marketable. My opinion is that if no one questions the methods, there can be no improvement. Consider all the advancements.

The dowelling method is the method I'll probably go with, dowelling before profiling due to existing tooling.

Someone I talked to years ago split his doors in half and laminated the piece back using an insert type tenon method. Trying to explain this method properly from memory. The tenon was dog-bone shaped and laid into a pocket that he had jig routed out and pumped with glue. This seemed to be more labor intensive than the old fashioned way. He also explained that it helped the warping. My question about the glue with just the cope was because of the technology of the various adhesives (i.e. the weld is stronger that the steel?).

From contributor S:
I saw a long time ago a study on the strength of joints, full m&t being strongest, floating tenon/spline next, then dowels, then pocket screws, last simple butt joint. Of course, all dependant on good execution. The looser the joint, the weaker. In my own opinion, the biggest force working against a door is gravity, and a nice tall bottom rail and lock rail with lots of glue surface are the best combatants.

From contributor J:
I make 1.75" , 2" , or 2.25" stave core doors with 1/4" or greater skins. I simply assemble them on a clamp rack with Titebond 2, and have never encountered any problems. Like anyone worth their weight in sawdust, I'm always looking to improve my methods. If I could implement a full M&T into the construction, without excessive cost (a rather subjective term, eh!), I'd love to offer that as an option. Fortunately, my clientele aren't nearly as concerned with price as most markets.

So, would someone explain the "right" way to make a M&T rail and stile union within the cope? I have a Powermatic mortise machine, and also a Leigh MMT jig. I'm guessing that the tongue of the cope in the rail ends becomes the tenon, and that the receiving groove at the union gets mortised deep to receive that tenon?

This would require the use of a window mullion type shaper setup (stub spindle) and two passes with the coping sled, right?

From contributor W:
A stub spindle would be one way to make tenons efficiently, though in my experience not common. I can't name any sources for such a cutter. Another efficient way to make tenons on a large shaper is with a large-diameter coping head (mine are 9-inch) such as those made by Charles Schmidt. They use small knives with corrugated backs so the knives themselves are not expensive to have ground. This is the same tooling used on single-end tenoners. For mortises I like hollow-chisel or (much more expensive) oscillating-chisel machines, both of which avoid the whip problems associated with rotating cutters of routers or slot-mortisers plunging in 2 or 3 inches.

Contributor J, your assumptions are generally correct about placement of tenon and mortise. Just curious - your stave-core doors are stile-and-rail doors joined with glue andů what? Surely not a butt joint, which is mostly end grain?

On adhesives, I don't think there have been any recent advances that pertain to this question, nor will there ever be any. Epoxy has been around for decades and is about as waterproof in non-marine applications and as strong as you could want. There is never a point to having glue that is stronger than the wood itself, which is pretty easy to achieve. Much more important is resistance to wood movement, mechanical stress, and moisture. These factors, particularly the first, will eventually cause any glue to fail or at least lose part of its adhesion. So you want to design a joint that will provide strength even as the glue is starting to fail, the main reason why mortise and tenon is preferred. It also provides the best mechanical strength, lessening stress on the glue bond. An old English book I have refers to the regional variation of joinery in which doors were never glued, but held together with wedged mortise-and-tenons.

From contributor H:
I believe true M&T joinery is the traditional way to go, but if you want to use M&T with a decorative edge profile around the panels, then you need fairly large cutters. Instead I use cope and stick joinery combined with loose tenons between the rails/stiles, and I cut these in separate steps with fairly small equipment. The loose tenons provide the structural strength that is needed, while the cope and stick supplies the decorative edge profile around each panel. You could substitute dowels (or some other type of structural reinforcement) for the loose tenons, but in my opinion, cope and stick alone would not provide enough strength for a door.

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor C:
Freud has cutter set fs99-302 that would satisfy the problem of a longer stub tenon. The cutter set has a removable top cutter that allows for a longer tenon to be made. At least the concept would give those that want to use this method an idea of how it works and may be able to improvise using existing cutters in their shop.