The Perfect Joint

Woodworking magazines are a wealth of information. I’ve been a subscriber to Popular Woodworking since 1998, and I have plenty of editions of Fine Woodworking, Wood, Woodcraft, Shop Notes, Woodsmith and many others on my book shelf. Every so often I trot out an entire year’s collection and look through all of the offerings. Great step-by-step projects. Excellent technique articles. Outstanding tool reviews. Each edition has much to offer.

Of course, there are the times when you’ll seem to read the same article over and over again in several editions of different magazines. “Master the Mortise and Tenon!” “Mortise and Tenons made easy.” “Cut spot on mortises and tenons.” And, those may be in three different magazines in the span of eight months. Hey, let’s face it; some topics are so essential to woodworking that you need to see them every so often.

A few years ago, there was this big push to do an article that really left me confounded. I saw it first in Popular Woodworking, then later in Wood. I’m sure it must have been in a few others. It was a wood joint torture test. The premise of these articles is a sound one – let’s find out just how much strength do each of these joints have. It’s a topic that has reared its ugly head in woodworking circles for generations. The tests were devilishly simple, and looked like they were cooked up by Wile E. Coyote of Looney Tunes fame. Basically, a sample of each of frequently used joints – dovetails, mortise and tenon, rabbets, biscuits, etc. – were placed on a work platform, and a 55 pound anvil was dropped a certain distance onto the joint.

The results, as they say, were spectacular. All of these joints, save the beefiest mortise and tenon, were demolished. Splintered. Crushed. The test did expose some of the strengths and weaknesses of each type of joint. For instance, a 90 degree box joint totally flattened out after its date with the anvil, while the dovetail did offer some resistance to the force of gravity.

My only question, however, is what exactly was the point of each of these tests? Sure, if you routinely abuse your woodworking projects, you had better build them like timber framed structures. But, for a hope chest, is a box joint necessarily worse than dovetails? The tests also showed that wood breaks away from pocket screwed joints when hit by an anvil. If this makes you believe that a mortised and tenoned face frame is superior to one that’s pocket screwed, and you have the time and determination to do all of that extra work, be my guest. However, aren’t most face frames joined to the cabinet box, gaining tremendous support from the case itself? In some cases, speed of assembly AND strength are the guiding principles.

Some woodworkers may use this test as further evidence that the ‘Old Masters’ who built furniture before us used only classically cut joints because that is the only way to do it right. But, hey, if the Shakers were so creative as to create work saving devices as the circular saw blade and the washing machine, don’t you think they would have relished the opportunity to use pocket screws or biscuits if they had the opportunity?

Am I saying that there’s no place for classic woodworking joints? Not at all.

Am I saying that there’s no place for new-fangled woodworking joinery techniques? Absolutely not.

What I am saying is that woodworking is an incredible craft, and there is an incredible array of choices available to today’s woodworker. Feel free to pick the right joint for the right job, regardless of the era from which it came, or how it fared when an anvil fell on it.

11 thoughts on “The Perfect Joint”

  1. Great stuff Tom.

    One of the things I find to be so wonderful about shop time is the mixture of precision and choice. A tight fitting joint is great, regardless of the flavor.

    There are tons of techniques to choose from. That is one aspect to your project.

    The quality of the execution is another factor.

    We aren’t building snap together models from plastic parts that are the same size (though I understand some people find this fun). We are building what we believe is a quality piece of furniture for our family / client. The pallet is broad.

  2. OK, Skee, here’s a challenge…

    I tell you to build a drawer. Now, let’s see how many options for joints (Front joints and back joints, if you want to be picky) you can choose to build the drawer.

    I’ll go first:

    Through Dovetails.

  3. Coming from and bridge engineering/construction background I really have to weigh in on this. The tests as described are only a guideline. “Which joint will fail least spectacularly?” is the question that they’re trying to answer. Seeing and understanding that a box joint in inherently weaker than a mortise and tenon gives us amateurs a base to choose the appropriate construction techniques can make up for the joint’s weaknesses. Granted, for most of our projects, we don’t care about the strongest joint. The aesthetic/time trade off matter most.

  4. I guess design only calls for a certain amount of acceptable strength. After all, your only storing books on that bookshelf or a TV in the entertainment center. Why build the joint stronger and to exceed what the recommended service is for. That only adds more cost? Yes, no, maybe? Hey says the woodworker… my joint is fine! It’s just the cheap crap stuff, wood, material around the joint that failed… it wasn’t my fault!!!

    Hell, don’t ever try to move the piece to clean the carpets or paint the walls. You silly furniture owner… today’s furniture can’t even survive being loaded and unloaded from a moving truck without splintering, loosing or coming apart. What are you thinking…?

    What are acceptable strengths…? Stay tuned next week… we got a good one for ya folks, were gonna build an entire dining-room set using double sided tape.

  5. I agree – if you want a strong cabinet that will take an anvil assult then get one made of steel. Otherwise, fit the joinery to the need and style and try not to drop an anvil on your blanket chest.

    However, I made a small, half-sized shed for my mother’s garbage cans and joined the door frames with pocket screws. as you might expect, the door frames fell apart within 4 months and three windy storms. Granted, I’m not a very good woodworker and never had experience with M&T joints, so pocket screws seemed like a better test of my abilities at that time. Now, the remainder of the door frames sit in my garage, awaiting new, M&T built frames.

    I guess I can see that an article (or slew of articles) to showcase how strong certain joints are can be very helpful to a newbie like myself, but I do agree that the obvious choice for the door frames would’ve been M&T all along. Anyone looking to make such a project would’ve came to that conclusion with minimal research and without an anvil test. I should’ve done it from the beginning, but I didn’t want to test out a “new” technique (new to me, that is) on someone else’s project………of course, I screwed it up anyway.

    Even though we would like to see woodworking from a fresh perspective on a regular basis, it seems that new techniques are few and far between. Those guys got to sell magazines, and they have to find stuff to put in them, after all 🙂


  6. I don’t know that I fully agree with Ace on this one.

    ToddT makes a good point — the joint “torture” tests really only answer the question of which joints fail the least spectacularly. Another way of looking at it is asking how much safety margin a given joint provides to the task.

    The disconnect is what’s the task? Ace hints at that — we don’t have corresponding usage guidelines of what the piece is likely to experience in it’s lifetime. Nobody has a good rule of thumb for the ft-lbs of force my toddler is going to deliver to his dresser when he climbs up on there and jumps up and down. Or how much stress is going to be applied when he pulls the drawers out and uses those as steps to get on top of the dresser. Thank god for those 100lbs slides! Granted, I can figure these things out thanks to my aging and fading engineering skills, but the number of possibilites is quite large.

    I’ll readily agree that the infamous anvil test was kind of hokey. Ok, not even kind of. It was. And even the contributing magazine’s editors agreed that it was hokey. But later tests have gotten a lot more rigorous and are using the correct equipment to test those loads. This gives us relative orders of magnitude.

    So what does having a safety margin (or rough guess thereof) in a relative order of magnitude tell us? We can feed that in our decision process when we’re trying to figure how we’re going to build it and what it will cost. It’s a relative piece of information that we use to determine end “cost”. Cost in the sense of time involved as well as monetary cost. Engineers recognize that there are always some sort of constraints on any project. Engineering is really about picking where you hope to end up on the constraint curves, balancing time, financial cost, and reliability amongs others.

    I can take that knowledge, add in other relative knowledge to make a final determination. Such as the time I was carrying half of the drawer to the workbench, tripped, fell, slammed myself into the pointy end where the two boards met and say “gee, that really should have held up better. I should use a better joint when I rebuild this.”

    I think the mian point of Tom’s article was that we have to examine new information and technologies as they become available. Few people still use an out-house for their primary toilet. Even fewer folk are arguing that we should return to those days. Sure, the outhouse looks rustic and might evoke delusional memories of “better times.” All of that quickly fades when you get within scent of the outhouse though.

    Great article, Tom! Keep ’em coming.

  7. Glen, perhaps I can cast a raking light over my previous post.

    My memory of most of the joint strength test’s usually illustrate pull-apart and shear failures. In addition, joints can fail from a torque or twisting action.

    I would prefer to see that we as woodworkers, plan for the worst. Toddlers do climb on things if were not paying attention. Wouldn’t it be nice not to have the stupid piece break if junior did get up on it? We as woodworkers should do the best work possible. Use quality joinery technique. We should desire to distance ourselves from production furniture, don’t we? As Engineers do make new things and make old things better. Granted that, its up to us as woodworkers to have the final say. And why no beat our chest that our joinery strength is superior to the adjacent wood area of the joint. We do live in a cover your a** society.

    I would prefer to have a nice stiff strong and supported joint. That won’t loosen and can withstand torque like the rigors of wrangling a dresser over the carpet to get my sock that fell behind it, as well as, clean behind my dresser once yearly, and that be in springtime.


  8. Ace, you clean behind your dresser once a year? What are you, a neat freak? 😀

    I’m sure if you looked behind my dresser, you’d find Jimmy Hoffa held hostage by a herd of rampaging dust bunnies…

  9. Reminds me of the never ending car tests. Yes, I am so happy that a car can get up to 190 mph, but when in my lifetime will I ever drive the thing that fast?!?! Just an example of useless statistics. It would be nice if there was an applicable test performed. I would be interested to see how various joint used on a chair would hold up, testing various forces like, weight on the seat, rocking motion resistance, etc… It would be nice to know if other non-traditional joints for chairs would work, as an example.
    Anywho, I am thankful they are still testing machines practically. I am not sure if I need to know that my SCMS can survive a 3 story fall or if my scroll saw can get to 176 mph and stop in 30 ft.

  10. I’m relatively new to woodworking. In order to easily make joints I bought a gadget called the WoodRat. It employs a router and works much like a milling machine to crank out box joints, m&t, dovetails and bridle joints. Because of my career in aviation, I thought those destructive test reports were fascinating.

    I am sitting in a kitchen that was completely remodeled exactly 13 years ago. I compare the condition of the cabinets in our brand new house in Northern California and I can see signs of distress in the joints. In particular, drawers tend to get hammered out of shape by their heavy contents. A simple drawer lock edge is not as strong as a dovetail. I’ve got both and can see the strength advantage of the DT.

    And doors where the rail and stile are bonded with a simple cope & stick edge are not as tight as ones with a true mortise and tenon. But I must say I thought that most of the arguments put forth in those test results apply to furniture you sit in. I would use biscuits happily in most of my work, but only a true M&T for a chair.

    Gary Curtis

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.