FWW Joint Strength Test

[Chris Padilla]

So this should be a doozy of a thread in which to discuss the findings that FWW did on 18 different ways to join, basically, a stile and rail.

The half-lap joint won for the strongest joint.

It was VERY interesting how/where the various joints failed.

I don't have the mag in front of me but I will later today and I can fill in more infor if someone else doesn't.

Dowelmax beat Domino but trad M&T beat both of them and I forget where the floating tenon came in....

[Clifford Mescher]

So this should be a doozy of a thread in which to discuss the findings that FWW did on 18 different ways to join, basically, a stile and rail.

The half-lap joint won for the strongest joint.

It was VERY interesting how/where the various joints failed.

I don't have the mag in front of me but I will later today and I can fill in more infor if someone else doesn't.

Dowelmax beat Domino but trad M&T beat both of them and I forget where the floating tenon came in....
What issue are we talking about Chris? Maybe it will be in the mail today? Clifford

[Doug Shepard]

The thing that surprised me was the bridal joint. They didn't include that on their last test, and I assumed it might come out a bit ahead of the half lap due to more glue contact. Instead it comes in just slightly behind the half lap. Obviously the wood thickness of the joint parts must be playing a part here too.

[Chris Padilla]

What issue are we talking about Chris? Maybe it will be in the mail today? Clifford
Yes, mine arrived in the mail last night.

[Chris Padilla]

The thing that surprised me was the bridal joint. They didn't include that on their last test, and I assumed it might come out a bit ahead of the half lap due to more glue contact. Instead it comes in just slightly behind the half lap. Obviously the wood thickness of the joint parts must be playing a part here too.

Yes, wood thickness and glue surface area are the two items that stand out for maximizing the strength of a joint for the specific racking test they were doing.

[Clifford Mescher]

Yes, mine arrived in the mail last night.
Yep, it just arrived in the mail. Interesting. Dowel Max beat Domino but suspect that Festooligans will complain that they used only one domino. Clifford.

[Chris Padilla]

We had to pull the thread but Jim Lindsay, Dowelmax inventor, had some issues with the testing.

I need to go back and read more carefully everything they did and calculate glue surface area and such to get a better comparison feel but I thought FWW did a good job of TRYING to make fair comparisons.

One thing they need to touch on more is WHAT is an appropriate strength. What does surviving 1000 lbs/area pressure buy you over "only" surviving 500 lbs/area pressure? Is it much? Does it depend? Is 500 good enough?

Since the trad M&T is pretty well time-tested, I think using its numbers is a good place to start. Old chairs rule here....

[John Keeton]

Here is what I took from this article - nearly all of the joints held where there was contact. So, the Domino joint failed at the termination point of the Domino, the dowel joint failed at the termination point of the dowel, etc. In those situations, the stile fractured along the grain. I don't have the article in front of me either, but seems the wood was cherry. That being the case, it was the wood that failed - not the joint.

The test really is deceptive in that regard. If there is sufficient force on the wood to fracture it, then I don't know that this test has much practical application, except to say that some of the methods just simply do not have enough glue contact, i.e., stub tenon, cope/stick - for cherry or whatever wood was used. Oak may produce a different result.

Most of my projects are not built for use in a military environment. While I am now a Domino owner (gloat, gloat!!), I don't think this test effects my attitude toward any of the joints that I may use in various applications.

Ultimately, these findings should give us some valuable information to use in deciding which joints to use where. And, that is all. I am not going to start putting everything together with half-laps just because they are the strongest. That just is not practical, nor necessary.

[glenn bradley]

I always find these interesting as well. Obviously, part selection can vary these results so multiple tests of each joint help homogenize the result..

[Michael McCoy]

Yep, it just arrived in the mail. Interesting. Dowel Max beat Domino but suspect that Festooligans will complain that they used only one domino. Clifford.

I just bought a Domino so I have to assert that the test has to be flawed. ;)

[Chris Padilla]

I just bought a Domino so I have to assert that the test has to be flawed. ;)

Being a Domino owner, I was thinking as I read the mag that Festool needs to come out with a larger/longer domino and associated cutter! :)

[Greg Peterson]

Festooligans! I love it! Good one Cliff.

[Peter Quinn]

...it was the wood that failed - not the joint.
The test really is deceptive in that regard. .


Funny thing, I was logging in to see if anybody else had read this article, just got my copy last night and was reading it at lunch. My impressions are similar to the exerts listed above from John's insightful thoughts.

I find these tests nearly meaningless and banal at best. What I see on careful examination of the pictures is a series of joints that held fast and a series of long grain fractures on the edge of various pieces of cherry. I am not quick to jump to any conclusions about each joints average strength, nor am I ready to change any methods I use or purchase new tools or lose old ones based on this testing method. Though I am strangely pleased that my old friends the half lap and bridle joint scored so well on the test anyway; these were the first joints I learned to make after all.

Do they assume that every piece of cherry has the same relative strength to begin with? Look at the pictures, look at the thickness of the annual ring spacing and ask your self "Is it possible that the weakest cuts of wood are the first to fail irrespective of the joinery involved?" This is after all a natural product with gross variations of strength over its range, not a standardized fabrication like steel or aluminum. Or is the suggestion that the deeper into the edge of a style a rail joint penetrates the greater the strength it will provide before the inevitable wood failure?

I further wonder how each joint may sustain a 'shock' force rather than a long slow crushing as created by the servo-hydraulic testing machine. Suppose a portly friend sits in a chair you have crafted. Will he gently lower himself down, adding weight incrementally? Or will he plop his but down hard and take a load off? Does it make a difference? How about child swinging from a cabinet door? (not that I ever did this as a child mind you...)

Wood being a naturally flexible product seems able to bend a bit as force is applied but some wood seems to break more easily when force is applied in a quick sharp manner. Ever break a sticker or thin cutoff over your knee? Bend it and it will bend considerably. Smack it on the edge of a stout table, and it will break. The editors in FWW do suggest in a sidebar that the test results are not indicative of a joints potential to withstand the stress of seasonal movement and use over time. Not sure how to simulate these factors in a lab, but without taking them into consideration what use are these tests really?

[Chuck Tringo]

I got it last night as well and agree ther are some flaws...one I saw was that for a pretty wide joint, it looks like 1.5 inch dowels were used instead of 2 inch, and yes a second domino would easily fit in the joint they showed. If I did use my dowelmax (ninja gloat) for such a wide joint I would use the longest dowels I had available, maybe even make some of my own from dowel rods to be long enough. Or maybe Jim needs to find a supplier who can make compressed dowels in 2.5 to 3 inch lengths :D

[Chris Padilla]

Perhaps it is splitting hairs decide if the wood broke or the "joint" broke. To have a joint, it involves putting together two or more pieces of wood. If the joint fails, it fails: be it the glue failed or the wood failed.

Since they used cherry, that kinda sorta cancels out through all the tests. Yes, being a natual product, no two pieces are EXACTLY alike so you have to take it with some grain of salt so add a tolerance of +/- 10% (or whatever) to all the numbers read of the test machine.

LMOL neglected to bring in my mag for lunchtime reading so I cannot comment much further on the details of each joint, size, glue area, etc. of each joint to see if they matched them up well or not.

Ultimately, and I'm drawing similar conclustions as everyone else I think: it comes down to glue area and wood "thickness" per se. Maximize this within the joint and it appears that it will be stronger according to this testing method.

It is an interesting test and it would be further interesting to see other tests like the shock one that Peter suggested.

[Ben West]

I've not yet seen this article, but should soon.

In any event, there have been several of these type tests published the past few years.

In general, be careful interpretting these results. Rarely do magazines implement testing prototcols that are rigorous enough to account for variances in the wood and in assembly procedures. In reality, one would have to duplicate each joint many times to account for these variances.

This FWW article may have done a good job, but that is not the norm with this kind of article.

In the end, does it really matter? I've seen very, very few pieces of furniture with failed joints because they weren't strong enough. It seems traditional M&T, domino, and dowelmax -- if applied correctly -- all produce joints strong enough for all practical furniture making.

[Chris Friesen]

With the bridle joint I'm not surprised that it came in weaker than the half-lap. The "single tenon" piece was only 1/4" thick. If I were making a bridle joint optimised for strength, I'd make the single tenon 3/8" thick, with each of the "double tenons" 3/16" thick. This would balance the amount of wood on each side of the joint, the same way that the thicker mortise-and-tenon joint does.

Their suggested "better way to spline a miter" basically turns the miter into a variation on the bridle joint.

The fact that the thicker tenon gives stronger joints is no surprise...I've seen many sources suggest that the tenon could be up to half the thickness of the piece being mortised, especially if the joint is machine made rather than being chiseled by hand. It's also not a surprise that the pinned/wedged joints were weaker. These additions are there primarily to keep the joint tight over time or to compensate for an initially sloppy fit, not to make it stronger overall.

[Peter Quinn]

I should add to my previous comments that on closer reading of the FWW stress test article I noticed they made and broke five samples of each joint, so the breaking force is an average of five samples which probably nullifies my suspicion that the strength of the individual pieces of wood in question is a strong factor. I still don't like these tests. Just a shame to waste all that beautiful wood when we all know (add your preferred method of joinery here) is the best if done correctly.:rolleyes::rolleyes::rolleyes:

Darn scientists. Perhaps they could bring in Pete Townshend to smash those joints guitar style and see how they survive my proposed shock test?

[Steven Bolton]

Was there anything about pocket screws? I like them, but I cannot believe they are as strong as mortise and tenon (as advertised).

Steve Bolton

[Peter Quinn]

Was there anything about pocket screws? I like them, but I cannot believe they are as strong as mortise and tenon (as advertised).

Steve Bolton

Pocket screws took 698# peak load before failure, 1/4" traditional M&T took 717# peak load before failure. Pretty close. A 3/8" tenon took 1475# before failure, considerably stronger. I for one will take a joint that can hold 700# any day. And a 3/4 ton door joint? That's just silly.

[Bruce Wrenn]

Only the butt, and miter joints failed. In all the rest, the wood failed, not the joint. Article could have been renamed " Wood With a Larger Cross Section is Stronger. " DUH?

[Mark Bolton]

I further wonder how each joint may sustain a 'shock' force rather than a long slow crushing as created by the servo-hydraulic testing machine. Suppose a portly friend sits in a chair you have crafted. Will he gently lower himself down, adding weight incrementally? Or will he plop his but down hard and take a load off? Does it make a difference? How about child swinging from a cabinet door? (not that I ever did this as a child mind you...)

I am new to sawmill and this is only my second post but...

As a manufacturer of wood products one can not include such variables in ones work. Operating in the most litigious and warranty laden environment in the world (US) one has to build work in hopes of withstanding the worst possible environment imaginable. More than likely its a situation you never would have imagined. There is no saying "well how hard did the guy sit on my chair when it broke?" As a manufacturer you are charged with the brutal task of foreseeing that possible eventuality.

The simple fact of the mater is we live in a world where someone today walks into a store, within a reasonable period of install, with a cabinet door broken down into its component parts and they are likely to be shipped a free replacement. It matters not whether they reveal that their 150lb kid was using it as a ladder to get to the reecees piecees or not. We as manufacturers in the market have no choice but to compete.

What strikes me as interesting is that standards remain true, simple = better. No matter how badly we all become obsessed with spending obscene amounts of money, a joint we have all been making for years with basic tools performs.

Mark

[Mike Henderson]

Only the butt, and miter joints failed. In all the rest, the wood failed, not the joint. Article could have been renamed " Wood With a Larger Cross Section is Stronger. " DUH?
That was my reaction when I read it. To me the test said, "Modern glues are stronger than wood, so make the joint hefty if you want the maximum strength."

The article was really about which technique has the most "mass" in the joinery, and is why the lap joint came out best.

Mike

[Randy Klein]

I like that they added a blurb discussing seasonal wood movement. That repeated cross-grain movement will eventually stress the joint. And that they acknowledged the test occurred right after the glue cured. It seems that they were expecting some fodder along those lines.

It'd be interesting to add a few decades of time, however unrealistic that may be, to the test and see how they hold up. I'd suspect that the top 2 - half lap and bridle joints wouldn't be the top 2 after that much time.

[Thomas Knighton]

I would imagine that they could put the joints in a chamber that increases and decreases humidity to such an extent that it would mimic a decade or so of time, but that would probably be cost prohibitive...if it truly is possible.

Tom

[David Keller NC]

"Since they used cherry, that kinda sorta cancels out through all the tests. Yes, being a natual product, no two pieces are EXACTLY alike so you have to take it with some grain of salt so add a tolerance of +/- 10% (or whatever) to all the numbers read of the test machine."

Actually, their choice of wood was extremely poor for the purpose for which they designed the test (testing the strength of various joint configurations). The ultimate conclusion of the test could've been phrased correctly that "different joint configurations varied the strength of the individual joint components - almost all of the joints failed where the wood was weakest". Their rather suspect conclusion that a bridle joint was the strongest because it had the most glue area is just flat out incorrect, based on their own test protocols and results.

What they should've done is realized that their wood choice was inappropriate based on the first few test joint failures. There are a lot of common, inexpensive species they could've chosen that would've been less brittle and would have likely allowed them to test the strength of the joint, not the wood. The ultimate one that comes to mind would've been hickory. Not exactly a common cabinet wood, but it would've allowed them to draw a conclusion about joint configurations.

This is, by the way, a common theme that runs throughout FWW's tests and reviews over the last few years. They do come up with pretty good article ideas, but their execution on those ideas is abyssmal. After reading some of the comments by some of the FWW editorial staff on Knots, this doesn't surprise me. They sorely lack appropriate leadership to focus their creativity and apply some logic to how they go about them.

[Greg Pavlov]

"Actually, their choice of wood was extremely poor for the purpose for which they designed the test (testing the strength of various joint configurations). The ultimate conclusion of the test could've been phrased correctly that "different joint configurations varied the strength of the individual joint components - almost all of the joints failed where the wood was weakest". Their rather suspect conclusion that a bridle joint was the strongest because it had the most glue area is just flat out incorrect, based on their own test protocols and results.

What they should've done is realized that their wood choice was inappropriate based on the first few test joint failures.
While you are probably right from a proper test approach point of view, the reality is that a lot more people use cherry for furniture rather than hickory, so the results will have more relevance for them.

[Joe Jensen]

These tests while interesting, don't seem to be of much value in the real world of furniture. I've made tons and tons of cabinet doors and cabinets over the past 30 years. Other than in the first few years, I only do but joints on flat glue ups, and I only use a cabinet set on the shaper for doors. I have never had a failure. I am meticulous on board prep, and fiend for straight flat boards.

[Clifford Mescher]

[quote=David Keller NC;996897figurations.

This is, by the way, a common theme that runs throughout FWW's tests and reviews over the last few years. They do come up with pretty good article ideas, but their execution on those ideas is abyssmal. After reading some of the comments by some of the FWW editorial staff on Knots, this doesn't surprise me. They sorely lack appropriate leadership to focus their creativity and apply some logic to how they go about them.
[/QUOTE]I couldn't let this go by without saying that I like reading Fine Woodworking magazine and I think they do an admirable job.Clifford.

[David Keller NC]

"I couldn't let this go by without saying that I like reading Fine Woodworking magazine and I think they do an admirable job.Clifford."

I still have a subscription, but those of us that have long-term subscriptions and have read the magazine since the 1980's realize that there's been a remarkable dimunition in quality of the magazine, both in the topics covered, the writing, and the completeness of the articles. There was a time when one would never see a tool review in the pages of FWW. While it's true that there's a place for tool reviews, FWW isn't that place.

There are several dozen titles on the subject of WW on the newstands, and it's highly inappropriate for FWW to attempt to dumb down their content in an effort to be more like those mags. There's a place for everything, including magazines filled with nothing but tool reviews and beginner projects, but that title is American Woodworker, not Fine Woodworking. There are many of us that have paid for subscriptions on a 5 year basis that are really, really not happy with the "new direction" of the magazine. It's a race to the bottom, IMO.

"While you are probably right from a proper test approach point of view, the reality is that a lot more people use cherry for furniture rather than hickory, so the results will have more relevance for them."

The problem is the labeling. A correct title would be "strength of cherry in various woodworking joints" not "strength of various glued woodworking joints". Because of this rather glaring ignoring of how the joints failed, the results do not translate to the various joint configurations tested in many other common cabinet woods.

[Joe Cunningham]

I forgot the issue at my office, so now I am snow-bound without a new FWW to read.

From my quick reading, my biggest take-away was that a thicker tenon seems to be strong in M&Ts. I wonder how a half-lapped miter would have done? I used that on a recent shop cabinet and found it pretty easy to cut and assemble (I use hand tools in probably 90% of my projects). Likely not a commonly used joint I guess.

I did find it a bit odd that pegged or wedged M&Ts didn't do as well as the same size M&T. And it did seem like the wood gave up the ghost before the joints in most instances.

[Clifford Mescher]

"I couldn't let this go by without saying that I like reading Fine Woodworking magazine and I think they do an admirable job.Clifford."

I still have a subscription, but those of us that have long-term subscriptions and have read the magazine since the 1980's realize that there's been a remarkable dimunition in quality of the magazine, both in the topics covered, the writing, and the completeness of the articles. There was a time when one would never see a tool review in the pages of FWW. While it's true that there's a place for tool reviews, FWW isn't that place.

There are several dozen titles on the subject of WW on the newstands, and it's highly inappropriate for FWW to attempt to dumb down their content in an effort to be more like those mags. There's a place for everything, including magazines filled with nothing but tool reviews and beginner projects, but that title is American Woodworker, not Fine Woodworking. There are many of us that have paid for subscriptions on a 5 year basis that are really, really not happy with the "new direction" of the magazine. It's a race to the bottom, IMO.

"While you are probably right from a proper test approach point of view, the reality is that a lot more people use cherry for furniture rather than hickory, so the results will have more relevance for them."

The problem is the labeling. A correct title would be "strength of cherry in various woodworking joints" not "strength of various glued woodworking joints". Because of this rather glaring ignoring of how the joints failed, the results do not translate to the various joint configurations tested in many other common cabinet woods.
I just came from my basement after looking through a few issues of early 90's FWW magazines. Yes, there is a sharp contrast between then and now. But to be fair, I have seen this "Dumbing-down" in all forms of media. This has actually been happening for at least 25 years in our educational institutions. They say a high school graduate of the early 60's is equivalent to a 3rd year college student of today.
I will not venture into the quality of movies that are made today except to say that I just started to receive the Turner Classic Movie channel and there is a stark difference in the sophistication of the dialogue of yesteryear.
In closing, the magazines and everything else today "is what it is". You have to remember that the people who published the magazine in the 80's were of a different era and unfortunetly they are dead or retired and they are not in charge anymore. Clifford

[David Keller NC]

"In closing, the magazines and everything else today "is what it is". You have to remember that the people who published the magazine in the 80's were of a different era and unfortunetly they are dead or retired and they are not in charge anymore."

True enough. I believe that the daughter of the founder took over a couple of years ago. Unfortunately, though, it appears that some consulting by "lean sigma" idiots has taken place. Several of the editors have said some remarkably stupid things in posts on Knots defending the shift in focus of the magazine. In particular, they note that their "surveys say this is the kind of content the market prefers". The idiocy of running your business without a quality objective and an empahsis on marketing surveys results in things like "New Coke" - one of the biggest disasters in modern business history.

Moreover, those of us with subscriptions that last well into the 'teens are forced to sit by and watch the magazine be trashed. It's rather doubtful we could get a meaningful refund on the remaining subscription issues (but of course that's not the point - the point is having to watch a beloved institution go down the tubes.).

[Clifford Mescher]

In particular, they note that their "surveys say this is the kind of content the market prefers". The idiocy of running your business without a quality objective and an empahsis on marketing surveys results in things like "New Coke" - one of the biggest disasters in modern business history.


Exactly. Their intent is to increase circulation. "Quality objective" Now isn't that a quaint idea? Clifford

[Peter Quinn]

I have a nearly complete subscription of FWW going back to the 70's, so if you just started reading in the 80's you are Johnny come lately and missed a lot of good content!:D The subscription belonged to a defunct wood carving business in the condo park I work in, mine for the taking after they sold the building and its contents to my boss. I am not old enough to have been subscribing back then. Saved it from the dumpster! Sort of represents a sociologic survey of the changes in the wood working community over the past 35 years.

Did you know it used to be all black and white? I could publish a better looking magazine on my computer at this point. And yes, way back in the 70's they had tool reviews, I am looking at one on the RAS and another on 15" planers! Not the glossy consumer reports tool shoot outs with numbers trying to quantify subjective information you see now, but reviews none the less. I could do without those in the pages of the regular magazine and would prefer they keep them in a separate annual publication for those who like that sort of thing, in case anyone from FWW is listening or cares what I think. Perhaps you could call it "Tool Junkie Issue: Number crunching beyond reason". I just might buy it but don't want it littering the pages of the general subscription.

Having read the whole magazine from year one (1974?) I can say it has certainly changed. There is still plenty of fine content to be had, just less of it and less often. It seems in the 70's America was riding a wave of resurgence in wood working and general artisanship popularized by boomers and hippies from the 60's and beats from the 1950's that wanted a return to honest living working with their hands. Many of these people left other careers, sort of dropped out to take up their craft. Became professional wood workers. Nearly every page of those old FWW issues was a master class, nearly every article, unlike the scant few pages dedicated to the highest level in their pages today.

So their are still many fine wood working professionals today, maybe more than ever, making a living making things. But it seems the popular belief today is that the present moving force in wood working, at least by the numbers, is not the professional community, but those that do it as a hobby! Dedicated artisans of the highest level and aspiring craftsmen and women plying their art....for pleasure alone! And it seems each year in an effort to capture the more fledgling part of this community FWW prints articles like "Best out feed table Ever".

I am going to be honest here: No decent skilled wood worker I know, professional or otherwise, needs so many pages dedicated to something this banal. Just plane stupid. I miss the old articles like "Tage Frid veneers the most incredible side table you have ever seen". Not the actual title, just my synopsis of its content. So maybe some need this type of advice and I don't fault them for it. Perhaps the time has come to publish two magazines, FWW and a spin off for the less advanced? Don't know if that would be financially viable. In the mean time, I agree with others that for a variety of reasons not all of which I claim to understand, the "FINE" in FWW has become a bit thin and tarnished of late relative to its beginnings. I still subscribe as they are IMO still the best publication on the subject I can find.

[Jim Lindsay]

I have reviewed the FWW article and have come to the conclusion that it is not just seriously flawed, but seriously undermines the intellect of the average reader. All other things being equal, i.e. curing period, glue type, wood type and dimensions, two entities exist which could fail during the stress process – (1) the joint and (2) the wood. An inquisitive reader will undoubtedly ask if the dowel joint did not fail and the wood failed at 759 lbs., what magical ingredient exists with respect to the mortise and tenon joint which would enable that particular joint to fail at 1,444 lbs. The answer is two-fold and fundamental, i.e. size and orientation.

Size: Referring to the 3/8 mortise and tenon, (FWW page 39), the size of the tenon pictured is at least 1 ¾ inches long, which is virtually twice the effective length of the 3/8 dowel which penetrates only 1 inch into the work piece.

Orientation: It will be seen that the tenon is intentionally positioned away from the top edge, well away from the area under tension and positioned close to the area under compression. No such preference was provided for the dowel joint.

I have a prediction and a challenge for FWW. Fashion a M & T joint which is equal in length to the penetration of the dowel and position the joint so that the orientation is similar to that of the dowel joint. I can virtually guarantee that if the M & T joint does not fail first, the wood will fail at 759 lbs.

A first year engineering student at university would know that in order to conduct a comparative test, it is essential to compare apples to apples and not apples to oranges. To compare a #0 biscuit – 1.46 square inches surface area (penetration) with a larger than normal M & T, 8.12 square inches surface area (penetration), is nothing short of ludicrous.

[Rick Potter]

Let's settle this strength question. Here's the deal.

Send me cabinet doors with your choice of joints. For a small fee, I furnish kids of all sizes to climb, twist, slam, and run toys into them. After a few hours we can finally determine which ones hold up in the real world.

On another note, I have recently planted 22 trees in the yard, and therefore have some extra carbon credits to sell to anyone wishing to reduce their carbon footprint.

Rick Potter

[Chris Friesen]

I have a prediction and a challenge for FWW. Fashion a M & T joint which is equal in length to the penetration of the dowel and position the joint so that the orientation is similar to that of the dowel joint. I can virtually guarantee that if the M & T joint does not fail first, the wood will fail at 759 lbs.

The article actually discusses this topic, suggesting that the craftsperson make/choose longer dowels/dominos/beadlocks/etc.

I suspect these two joints were probably typical joints of their type. When making an M&T joint, the incremental cost of making it longer is minimal, so it's more common to size the joint appropriately.

With a dowel joint, the maker is likely using that technique it for speed and convenience, which would often mean using premade dowels. In the case of the test, they used the manufacturer's dowels. In any case, the failure mode of the dowels was to split the stile at the dowel tips rather than to blow out through the end grain, so I don't think the fact that one dowel was placed near the end had any impact on the failure mode.


To compare a #0 biscuit – 1.46 square inches surface area (penetration) with a larger than normal M & T, 8.12 square inches surface area (penetration), is nothing short of ludicrous.

I disagree. I think it makes sense to compare typical joints of each type in order to get an idea of how they rank.

The biscuit was at a disadvantage due to the dimensions of the sample. The slot for a #10 biscuit is 2 3/8" in length, so the #0 is the only one that would fit. They could probably have managed two biscuits though, which would have given twice the glue area.

I'm not sure why you call the M&T larger than normal. Their 3/8" tenon joint looks just about exactly like what I would use in a similar situation. One rule of thumb is a tenon length about 5x the thickness, or in this case almost 2". Given that the pieces are 2 1/2" wide, their sample looks about right.

[David Giles]

All testing has contains some relevant information. I came away with the following conclusions.

Almost all joinery methods are "good enough" for a typical application.

Thicker tenons are stronger. 5/16 to 3/8 tenons are better than 1/4" thick.

Longer tenons (or equivalent dowel or Domino) increase strength.

Pocket hole joints were surprisingly strong and can't be beat for fast and easy (utilitarian drawer boxes).

[Douglas Robinson]

I agree David!!

I am surprised that no one has noticed the parallel to the August 2007 issue about glue tests. FWW even went to the same place! This is recycling the same stuff year to year. I have heard that there used to be about a 4 year cycle of repetition. But this is ridiculous. I have EVERY issue of FWW, but when my subscription runs out I will not renew unless the quality improves between now and then.

"There's been a remarkable dimunition in quality of the magazine, both in the topics covered, the writing, and the completeness of the articles."

Not only that, but in NC we have had a lively discussion of how BAD the latest tools and shops issue was! No shop review, a crumby epoxy floor painting article, hand-tool jigs, issues with basement shops, (WHich had been covered better in a prvious Tools a& Shops issue) and the "Ultimate" outfeed table. Which was far from it.

"There are several dozen titles on the subject of WW on the newstands, and it's highly inappropriate for FWW to attempt to dumb down their content in an effort to be more like those mags. There's a place for everything, including magazines filled with nothing but tool reviews and beginner projects, but that title is American Woodworker, not Fine Woodworking. There are many of us that have paid for subscriptions on a 5 year basis that are really, really not happy with the "new direction" of the magazine. It's a race to the bottom, IMO."

YES ABSOLUTELY!!!! I really resent this. If FWW can not add to my store of woodworking knowledge, then I will not buy it!

Doug

[John Keeton]

All testing has contains some relevant information. I came away with the following conclusions.

Almost all joinery methods are "good enough" for a typical application.

Thicker tenons are stronger. 5/16 to 3/8 tenons are better than 1/4" thick.

Longer tenons (or equivalent dowel or Domino) increase strength.

Pocket hole joints were surprisingly strong and can't be beat for fast and easy (utilitarian drawer boxes).
But, didn't we all know this without FWW having to tell us??

[Chris Padilla]

But, didn't we all know this without FWW having to tell us??

Not necessarily! :)

[Craig Reynolds]

One thing to remember is joint strength is not just glue area, it's what part of the wood is being glued. The reason mortise and loose tenons are stronger than dowels is more long-grain glue surface on the tenons. The results of these tests aren't surprising at all.

[Matthew Hills]

In wood, the joint is more than the glued surfaces -- there is also the issue of the wood itself and how the grain from the two is integrated. (This is separate from the "glue long grain to long grain" mantra)

As others have noted, many of the failures were splits along the grain of a piece, in response to the shearing load there.

An interesting case would have been if they had taken milled the final test piece out of a single block of wood, with no glue joint. Or even cut the shape out of plywood.

I think the test was interesting, even if they didn't really lay out the underlying reasons behind the results. And I do appreciate their nod to some of the tradeoffs (half-laps aren't for everyone! :-)

Matt

[Ben West]

Like many of you, I find little use in the joint strength article. They don't compare apples to apples, and thus the results are really meaningless. And, any of the M&T type joints (dowels, domino, loose tenons, real M&T) are strong enough for 99% of what we do, if properly applied.

Having said that, I still enjoy FWW and will continue to subscribe to the online site. Gary Rogowski is an awfully good craftsman and writer, and many of their other contributors are too. While each issue has some things I don't find useful, each issue also has some things that are new information for me.

Having said that, I've become awfully impressed with Chris Schwarz. His bench book is one of the best woodworking related books I've ever read, and that's led me to read some of his articles and blog postings. He is very talented. To that end, Popular Woodworking or Woodworking Magazine may be getting some $$$ from me this year too.

[wil wolski]

Being an owner/user of the Domino system for over a year now,I can say that this one of my favorite tools and has paid for itself several times over.
I have never had a joint fail with this system but I am careful use the correct
amont of domino(es) per joint.
The FWW article was surprising and does not really apply the Domino properly IMHO.
That being said, I'm not sure what the FWW test was really meant to prove, all of the joinery system mentioned have their own merits, and no one can handle all applications.

Wil

[Jim Lindsay]

Hello SMC,

We're in the process of completing a response to the FWW article discussed in this thread.

When completed this will be our third series of strength tests, and will compare multiple dowel joints having 1 inch and 2 inch dowel penetrations into each workpiece with a mortise and tenon joint having a 2 inch long tenon.

So far we have the videos for the two multiple dowel joints here for the 1 inch here

http://ca.video.yahoo.com/watch/5415216/14263627

and for the 2 inch penetration here

http://ca.video.yahoo.com/watch/5415255/14263676

The dowels used are high quality standard multi groove pins manufactured in the US. We will have the mortise and tenon joint test completed shortly. Whatever happens with that, it seems clear from the videos so far that the strength provided is well in excess of what is required in practice for a piece of furniture.

Our earlier second round of tests using wood joints with 1 inch dowel penetration and 1 inch tenon length are here

http://ca.video.yahoo.com/watch/5415338/14263959

Regards,
Mike, Dowelmax.

[Chris Friesen]

We're in the process of completing a response to the FWW article discussed in this thread.

I respect the strength of multiple dowels, but I'd like you to be fair when doing the M&T test and not undersize the joint for the sake of a more favourable comparison.

If I'm going to go through the trouble of making a M&T joint it's because I'm trying to make it durable. I'm going to use the longest thickest widest tenon that I can. My rule of thumb is that the mortise should be 1/3 to 1/2 the thickness of the piece being mortised, and where possible the length of the tenon should be at least 5x it's thickness. The tenon will be as wide as possible while still allowing suitable shoulders (and a strong enough end when joining the corner of a frame).

[Mike Henderson]

I doubt if some joint strength tests are going to make a lot of difference.

First, most of the joint techniques are strong enough.

Second, some people choose their joinery technique for reasons beyond strength. For example, traditionalists will go with M&T because it's traditional. Someone else may choose Domino because it's fast.

Third, many (most?) people look askance at strength tests because they're all biased in one way or another. Any strength test by a product supplier is especially suspect because they'd hardly report unfavorable results, and are likely to design the test to favor their technique.

Mike

[Jim Lindsay]

I respect the strength of multiple dowels, but I'd like you to be fair when doing the M&T test and not undersize the joint for the sake of a more favourable comparison.

If I'm going to go through the trouble of making a M&T joint it's because I'm trying to make it durable. I'm going to use the longest thickest widest tenon that I can. My rule of thumb is that the mortise should be 1/3 to 1/2 the thickness of the piece being mortised, and where possible the length of the tenon should be at least 5x it's thickness. The tenon will be as wide as possible while still allowing suitable shoulders (and a strong enough end when joining the corner of a frame).

Chris,

We agree with you that the test pieces should have a comparable connection size. That is exactly our objection to the FWW test where a 2 inch long tenon was tested against a dowel joint with only 1 inch dowel penetration, and the Wood magazine test where a 1-3/4 wide x 1 inch long tenon was tested against only two 3/8 x 1-1/2 inch dowels. The Wood article declared the connections sizes at least, where the FWW article did not. Now that we have the video proof that the dowel joint with 2 inch dowel penetration is significantly stronger than the dowel joint with only 1 inch penetration, we can declare the FWW test invalid.

The tenons on the mortise and tenon joints we have for testing are 2-7/8 or 3 inches wide (to be confirmed) x 3/8 thick by 2 inches long, and are considerably better m&t joints than are typical. We're taking a bit of a risk coming out with this ahead of the test since we do not know what will happen. Whatever does happen with the m&t, the 1000 psi required to break the multiple dowel joint is already well in excess of what is practically required.

-Mike, Dowelmax.

[Jim Lindsay]

Third, many (most?) people look askance at strength tests because they're all biased in one way or another. Any strength test by a product supplier is especially suspect because they'd hardly report unfavorable results, and are likely to design the test to favor their technique.

Mike

The recent published strength tests by Wood magazine and FWW are clearly and demonstrably biased toward mortise and tenon joints. These are not tool suppliers or manufacturers as identified above.

In the July 2007 issue of Wood magazine, a m&t joint with 1-1/2" wide x 1" long tenon was tested against a dowel joint with only two 3/8" x 1-1/2" dowels. The combined width of the dowels was only 3/4" compared to 1-1/2" for the tenon. For a fair comparison, three 3/8" x 2" dowels should have been used.

In the February 2009 issue of FWW, a mortise and tenon joint with a 2" wide by 2" long tenon was tested against a dowel joint with only 1 inch of penetration into each workpiece. For a fair comparison, a dowel joint with two inches of penetration into each workpiece should have been used.

These are not small differences. Twice the width for the m&t in the Wood magazine case, and twice the length for the m&t in the FWW case.

In the April 2009 issue of FWW the magazine printed our objections to the test, but did not substantially address them. We have asked several times for a retest using longer dowels, and have been refused.

Our three series of strength tests have been done in response to the bias and procedural errors in the magazine tests, and have been conducted fairly. The results have be video'd and presented in several locations on the internet. The results are predictable and consistent. In the magazine tests, the dowel joints break the wood. In our tests, by adding a greater number of dowels, or increasing the length of the dowels, the force of the test is acting on a greater area of wood which is harder to break resulting in a stronger joint.

If there is a specific and factual argument as to why our strength tests are suspect, please provide this so that we can address it.

Mike, Dowelmax.

[Mike Henderson]

I spent my professional life in electronics and semiconductors. I've never seen a fair and unbiased product comparison put forward by a manufacturer. They simply would not publish any comparisons which did not favor their product.

I doubt if things are any different in the woodworking business.

Mike

[Cody Colston]

I read the joint test article again last night and after reading the responses here, there are some points that are obvious to me.


It matters now how "objective" a test is, someone will find fault with the methods, materials, tester, lab, etc. no matter if they have ever conducted any type of destructive test in their life.
Most of the responses are predictable, regardless of what test is being conducted; ie, is anyone surprised that Festooligans cried foul when their beloved joint was middle-of-the-pack...and barely ahead of the much-maligned biscuit (which also utilized only one).
Twenty years from now, readers of FWW will be bemoaning the fact that the mag has deteriorated in content from the great articles of the 2000's.
Is anyone going to change their joinery methods based on the results of the FWW article? I thought not. :D

[Brian Kent]

I
Is anyone going to change their joinery methods based on the results of the FWW article? I thought not. :D
[/LIST]

On one point I beg to differ. I'm still pretty new at this and I learned not to go too fancy - just use a joint that has plenty of long-grain glue area.;)

Otherwise, this has been a very entertaining thread with some good info too!

Brian

[David Keller NC]

"The recent published strength tests by Wood magazine and FWW are clearly and demonstrably biased toward mortise and tenon joints. These are not tool suppliers or manufacturers as identified above."

This is a resurrected thread, but a point (I think) I made earlier in the year bears repeating. Because the FWW test samples all failed (with one exception) in the wood, the test did not evaluate the strengths of different joints. What was evaluated was the strength of cherry in various geometrical configurations.

This may seem like a trivial difference, but it isn't. Certainly, if one only works in cherry, the results might have some practical, if narrow, application. But it is wholly incorrect and misleading to conclude that one joint type vs. another in the test is stronger.

In order to correctly test what the FWW trials were designed to do, the joint must fail along the glue line, which means that either a stronger, less brittle wood would have to be used (like hickory, for example), or a weaker glue.

I note this simply because FWW itself did not understand that they could not reach a valid conclusion about which joint was better/worse - only that cherry fails more readily in particular joint geometries than others.

[Ryan Stagg]

Twenty years from now, readers of FWW will be bemoaning the fact that the mag has deteriorated in content from the great articles of the 2000's.



Does that mean that in 20 years FWW will only be chocked full of articles about how to assemble Ikea furniture? :p

[Adam Cavaliere]

Does that mean that in 20 years FWW will only be chocked full of articles about how to assemble Ikea furniture? :p

Heck there are already people out there offering up their expertise in putting together Ikea furniture for a fee... I think there should be a trade magazine dedicated to them!

[John Keeton]

Here is what I took from this article - nearly all of the joints held where there was contact. So, the Domino joint failed at the termination point of the Domino, the dowel joint failed at the termination point of the dowel, etc. In those situations, the stile fractured along the grain. I don't have the article in front of me either, but seems the wood was cherry. That being the case, it was the wood that failed - not the joint.

The test really is deceptive in that regard. If there is sufficient force on the wood to fracture it, then I don't know that this test has much practical application, except to say that some of the methods just simply do not have enough glue contact, i.e., stub tenon, cope/stick - for cherry or whatever wood was used. Oak may produce a different result.

Most of my projects are not built for use in a military environment. While I am now a Domino owner (gloat, gloat!!), I don't think this test effects my attitude toward any of the joints that I may use in various applications.

Ultimately, these findings should give us some valuable information to use in deciding which joints to use where. And, that is all. I am not going to start putting everything together with half-laps just because they are the strongest. That just is not practical, nor necessary.And to reiterate what David has observed, this was my post early on regarding the FWW tests - they have limited value, as would any other joint tests outside of a strictly controlled scientific study.