Door Joinery: Haunched Tenons Versus Floating Tenons
In a loose tenon or spline joint, a careful attention to the fit of the mortise and tenon and a complete painting of both with appropriate glue will virtually eliminate failure in a lifetime.
I am a wood geek, so I often dissect older work (up to 100+ years old) to discover reasons for failure and for the opposite. I have removed doors from remodels that were more than 2" thick and over 100 years old that had loose tenon construction and were as solid as the day they were built. That includes doors that have been devoid of finish on the lower rail for years or have been flooded repeatedly. Conclusion? This joint works, was used quite often by hand craftsmen and was respected in the literature both then and now. I can make this door at a profit for far less than someone using a haunched joint.
The extended tongue joint is also a reasonably strong joint and the one I use for interior doors. My tenon is 1 1/2" long and represents a lot of glue area as well as a solid member to pin with a fastener from the bottom or top of the door. The only downside to this joint is the initial outlay for special tooling for your shapers that will run upwards of $1000 to start. Again, it has worked for several decades without fail for me. I even use it for my cabinet doors when I know they will get heavy use or I can get the payment for this type of construction. Remember, this entire debate is driven by what our customers are willing to pay for!
Selecting an appropriate adhesive for door construction is as important as the methodology of the construction itself. Using the old standard shop glue to make a 1 3/4" thick exterior door is simply a bad decision. A very small amount of research or a call to your adhesive product specialist will result in the discovery of far superior glues for this application. From the type III glue that I often use in the northeast to the marine grade epoxy that I used while managing a woodshop in the Caribbean, the choice will be dictated by your circumstances. Exterior doors need a waterproof glue. Interior doors should use a water resistant glue at the very least. If a form of that adhesive can be a long set type, even better.
I am of the "used to use solid wood and now make my components engineered" school. The simple and unavoidable truth is that the wood available today is of a much inferior grade in most cases than was available even 10 years ago. The use of smaller and smaller logs, and compression logs (trees that grew at an angle - something that is usually overlooked today - and will always move after milling, and continue to move) have convinced me that there is a major benefit to engineering your components, if done correctly.
First is the core. There is nothing inherently wrong with a finger jointed matrix, if the correct specie of wood is used (fir and yellow pine are my choices for their stability and strength) and the correct adhesive and manufacturing method are also used (deep grooves in the fingers - thermally activated epoxy glue). In fact, the alternating of grain direction that is a natural result of this method of manufacture are a benefit as it results in a part that is more resistant to warp and cupping. The thickness of the caps on the exposed and milled edges and the thickness of the thin sliced (by definition it is not veneer!) faces are of the utmost importance to the customer, as they usually focus on appearance, not construction.
My faces are never less than 1/4" thick and often 1/2" or better. This door is never going to have a problem with refinishing or sanding. Also, unless the adhesive manufacturer has had a failure in spec compliance for a batch of glue, I would say there is near zero likelihood of a failure of the adhesive bond on any part of the door.
So, with all this milling, gluing and attention to detail, why make a manufactured component at all? Quite simply put - it is a superior product if done correctly. The likelihood of this door warping, twisting or in any way moving because of exposure to extreme weather is small. It is easier and less expensive to make a door from solid wood. Period! The labor involved in making a door from engineered components far outstrips the cost of lumber in all but the most expensive woods like lacewood, zebrawood, ebony and such - all of which I have made with this method.
The single most effective finish for an exterior door in any climate is a teak or other heavy hardening oil that is applied every six months for the first few years and then yearly afterward for the life of the door. This will never peel, chip, crack or fail if the regime is adhered to - and therein lies the rub. Unless you have a very constant and close relationship with your customer and they are willing to pay for this service, you will need a compromise. Depending on the climate, exposure and other factors that the door is exposed to, you could choose a spar finish, an exterior poly, or another engineered exterior finish. Talk to your exterior coatings specialist in your area.
I was always of the mind, with my top of the line exterior doors, that I will send a craftsman/salesman around to my customers to recoat their doors as needed for the first several years. It is a service that opens eyes (and pocketbooks for additional work!) and assures that your doors always look incredible and acquire a deep lustrous look that is only attainable with a teak oil over time. It only takes one man an hour or two, even with protection work, and has the added benefit of convincing the customer how easy it is to do it himself (or to pay you to do it for him after the initial period!). Something to think about in this economy of hyper-competition in woodworking...
My customers know that if they have a problem with any door I make at any time in their lives, I will make it good. I have yet to need to do that for anyone.
From contributor L:
I'd like to second the notion of using loose or floating tenons. I'd challenge anyone to make a case that a solid tenon is any stronger. Add to this that the two mortises are infinitely more efficient to make, and in my opinion it's a no-brainer. The shoulder fit is always perfect since it's a simple cross cut. I should add you'll never forget to allow for tenons on both ends when cutting the rails to length. (Please don't ask.) Always nice to read a comment here that's right on. Contributor J nailed it.
Would you like to add information to this article?
Interested in writing or submitting an article?
Have a question about this article?
Have you reviewed the related Knowledge Base areas below?