Glue Myths: 2. Miters

[@gregthorne4292]

I think you hit the nail on the head with the “power of leverage” comment, and I think it’s where both of these myths stem from. End to end and miter glue ups aren’t inherently weak; they just happen to be the configurations that result in the most leverage placed on them since wood is typically cut the longest in the direction of the grain. That’s a hard story for master woodworkers to instill into apprentices over generations, though, so it got boiled down to dogma of “these joints are weak”.

[@johnnymartin2970]

"Conventional Wisdom" often gets misunderstood and misapplied by subsequent generations.

[@RenaissanceWW]

Love it "miters aren't weak, your joinery skill might be though" Well done sir.

[@ABaumstumpf]

That is the only conclusion that is really accurate here.

[@chriselliott726]

I think an important 'real world' factor is the inherent difficulty in clamping miters, even well cut ones. In many instances we use band clamps which do little more than hold the pieces in position rather than apply a good clamping pressure. Excellent video by the way. Thank you.

[@randsipe224]

I make a ton of frames. I typically use a frame made of threaded rod and corner braces to clamp the frame but I have also used the type of band clamp with the plastic corners. You probably know the type I’m referring to. In my experience both theses types can exert tremendous pressure on a frame. The reason is the multiplication of the force vectors in all 3 directions. As you may be aware a wedge can exert tremendous force. In a typical frame glue up each of the 8 miters are effectively acting as wedges multiplying the force of the clamp it’s self. The tension on the miters is many times greater than a typical straight clamp can exert. There are actually 3 vector forces on each miter. One horizontal, one vertical and one diagonal. A typical clamp base only one.

[@charlesenfield2192]

To the extent that ease of clamping affects the fit of the joint, I agree, but Patrick already addressed that. High clamping forces can be used to close up a poorly-fitted joint, and that's hard with miters. However, the real problem is the poorly fitted joint. Well cut miters require very little clamping force to achieve full strength. Holding them in position is all that's required. This is easy to test if you care to. Cut four small samples from the same board with straight, square edges. Glue them together in pairs. For one pair clamp with a rubber band, and the other squeeze it has hard as you like with the clamp of your choice. Once cured, break them both. I doubt you'll perceive a difference. This leads me to request a different test in the future. I've heard people suggest you can over-clamp, and "squeeze out all the glue", resulting in a weaker joint. I have no experience to suggest there's any truth to this one, nor given what I know about fluid mechanics, can I think of any way to actually squeeze out all the glue.

[@bobby-c7731]

I loved you in Schitts Creek

[@CeeJayThe13th]

I'm pretty sure that I've seen the myth of "clamping pressure" busted a few times here and there.

[@JasonPeltier]

@@bobby-c7731 my first thought, lol

[@fredbrookes7968]

Mr Sullivan, your series of careful, rigorous and appropriate tests is of enormous benefit to the woodworking community and industry. I offer my grateful thanks to you for making your knowledge available. With my best wishes, from UK.

[@CarterCathey]

I love it when you have a very specific question and you come to the internet and find somebody like you who has done such wonderful and precise research and then summarized it all in a 12 minute video for me. Really great work and great clarity of making the insights consumable by a layman.

[@paulferraby1202]

The lab test mode is very different to the sample 'frame' broken by hand. In the lab tests, the bending moment is across the weaker plane of the test pieces, i.e. you're bending across the thin section but with the mitre broken by hand, you're bending across the thick section. This means that wood section is inherently stronger in relation to the section of the joint (the wood is thicker, and the joint is thinner) and there is the added disadvantage of a stress raiser at the inner intersection that predetermines where the fracture will initiate -at the glued joint. in addition, the bending mode of a mitre bent that way is such that the fulcrum for the two opposing levers is on the outside of the joint and if you look at the resulting forces at the inside of teh joint, they will be orientated more along the grain axis rather than across the lignum. As an engineer, I very much appreciate your methodical approach and your focus on relative strength (and failure modes) using samples of identical geometry. I can't help feeling that comparing the effects of geometry would teach us more, but I guess that your test equipment wouldn’t be able to apply sufficient load to break the joint if you rotated those samples 90 degrees and you would need to scale the geometry down. Keep up the good work though. I enjoy the videos and the intelligent debate that stems from them.

[@richs5422]

This is a crucial point, thank you!

[@charlesenfield2192]

Very true, but probably irrelevant. Instead of loading a single miter, picture how miter joints are actually loaded in your projects. The shape of the final project is usually a closed loop of some sort, and the forces can't produce a large bending moment like in the sample he broke by hand. In practice, miter joints are subjected to tension and shear - the same forces produced in Patrick's test apparatus.

[@paulferraby1202]

@@charlesenfield2192I agree that it's unlikely that a mitre would be used in an 'open ended' way like that and that mitres are part of a closed structure. These joints are most stressed by loads that cause the joint to open or close such as loads on opposing corners or opposing loads along opposite sides. These cause tensile stresses at the inner corners of two opposing joints and the outer corners of the other two. My real point is not about mitres but about the effects of the joint geometry with respect to the width vs height of the glued section and the potential effect of stress raising corners at the interface. Patrick is suggesting that there are other aspects to consider in further studies, I'm proposing some.

[@bennettrogers7921]

I thought the same, the direction of forces exerted during his test are significantly different than his hands on demonstration. It would be tantamount to twisting the wood as opposed to prying. I appreciated his scientific approach and believe that there should be follow up research.

[@urablahblah]

Came here to say the same thing. Well, actually I came here to find someone else who said it better than I could and put an appreciative comment and thumbs up on their superior post.

[@billkurek5576]

Let’s not forget the impeccable setup and camera work plus easy to understand charts. I thoroughly enjoy your presentations.

[@AdamTheEvangelist24]

Well said Patrick. I appreciate your effort and demeanor. Recognizing that woodworkers are mostly intelligent people with a high standard of excellence is what makes the joy and challenges of woodworking so fulfilling. We appreciate what you are trying to do and thank you for your effort and spirit of excellence.

[@Bropann]

Thank you so much for your time and dedication to the art and craft of woodworking. You explode myths that have been dear to my heart for a long time and you do it with patience and clarity that easily dispels my previous assumptions. Kudos to you, Mr. Sullivan.

[@annaeliasson952]

"For those few who don't love vector calculus..." I for one have found vector calculus to be rather an acquired taste. Your videos, on the other hand, are easy to love.

[@robandsharonseddon-smith5216]

well of course, it's all about the rate of change of the thingamajig in the oriented direction of the whatsit, when it comes down to it. Simple, really....

[@MichaelOlsen-Engineer]

That's all well and good, but nobody is talking about vectoria claculus. After all, science is supposed to be gender neutral. ;-) LOL

[@supplanterjim]

@@MichaelOlsen-Engineer Well, this video is non-bindary.

[@JanStureNielsen]

As a physicist and woodworker, I love what you're doing with this series, Patrick Sullivan -- thank you for all the work you put into these so we all can learn more from your quantitative analysis.

[@supplanterjim]

It seems that a _lot_ of woodworkers have a strong background in math and engineering. Why _is_ that?

[@JanStureNielsen]

@@supplanterjim -- I suspect you're right that woodworkers skew STEM, perhaps because the technical barrier is sufficient to filter out many of the "crafts"-builders (

[@Keneo1]

@@supplanterjim it might seem that way because you are looking at anecdotal information and may have a bias based on your environment where you observe woodworkers. A bigger study with rigorous data collection that corrects for baseline and avoids common biases might reveal the full picture.

[@supplanterjim]

@@Keneo1 Ha ha ha! Was that _irony?_ Wonderful!

[@kylehadden7294]

I would greatly appreciate a video on spline strength!

[@leolandleo]

The amount of times I have referenced these glue myths videos since you released them is nutty. It might be decades before this myth stops being so widely believed.

[@perihelion7798]

Again, a long overdue video. Thanks for your efforts and your science.

[@bhoola123]

Thanks for making a 2nd episode on it. I did see some famous woodworker felt insulted to be proven wrong on this topic, but some others welcomed your efforts with open heart and claimed that they learned from it. Sir you did Great Job for woodworking community. We are very thankful. Hope we learn more from you in future.

[@kevindoyle4521]

I really like your approach. As a mechanical engineer with 40 years of experience designing and building manufacturing equipment and about he same amount of woodworking experience you hit all of the check marks. You gave a better approach than I could have.

[@bobfugazy4916]

My experience, and I'm a hobby woodworker, is that the more accurate you can make any woodworking joint, the better. Thank you for your time and effort in doing these analyses. We all need to learn how to better flatten and square our wood, myself included!

[@DaveParkerNZ66]

When I started making picture frames I did a very simple test by gluing one mitre joint together. After curing I could not break the joint with my hands. I reasoned that this would be strong enough for the purpose, particularly because the four joints each support each other. Since then I make all my picture frames by gluing the mitre joints. I haven't had a failure yet.

[@CeeJayThe13th]

Yeah you've gotta think about what you're building and what stresses are actually going to be on it. A picture frame doesn't need to be strong enough to dance on and a table doesn't need to be strong enough to have a dance contest on.

[@frickd7355]

As a retired engineer wood working for over 40 years I appreciate the method and style of presentation. It has been my experience that taking a strong joint with a poor design or poor construction will not work. Conversely a weal joint used in a proper design well constructed subject to proper forces could last for generations. Flaws in materials, steel or concrete or wood, will fail much sooner subject to design, construction and forces applied. Thank you for this series. I have also enjoyed the response from some well known woodworkers that this series has generated.

[@Hatchmade]

What about clamping force? I’d love to see a video on how clamping force affects glue bond strength. I’ve heard that over clamping can starve the joint from glue but I’ve never seen that myself.

[@mychalevenson7710]

I appreciate the addition of "simple joints." It helps clear up a lot of the issues from the previous video.

[@ABaumstumpf]

Would have been nice if he actually added any longgrain-joints. A Lapjoint really isn't hard to make yet many times stronger than endgrain joints.

[@SteveBakerIsHere]

Many thanks! An interesting result...but one I think needs to be better understood. One interesting (and fairly easy) experiment you can do is to use polarization to visualize the stress in most common plastics. If you take an L-shaped piece of acrylic, shine polarized light through it, and view it through another piece of polaroid - you can see the pretty rainbow patterns of stress within it. You should DEFINITELY do that - it's quite instructive! If you imagine a typical miter joint, the stress is MUCH greater on the inside corner than the outside. Interior corners concentrate stress PHENOMENALLY - which is why rounding those interior corners is a good idea when you can do it. When you try to pull open an actual miter joint by hand (as you do in the video) - the forces are building up selectively on the inside corner...and the outside corner is actually in compression. Glue can cope with HUGE amounts of compressive force. So applying forces in that way will cause the glue at the inside point to have MUCH more stress. Your test doesn't reproduce that. It applies the forces much more uniformly across the entire glued area. So you're right that the glue itself (when applied at 45 degrees to the grain) is stronger than on side grain and not as strong as as on end grain - but that misses the point about mitered joints and why they fail. So what's going awry here is that you're talking about glue strength - and the anti-miter people are talking about joint geometry strength and the way that forces within the joint are concentrated. A second problem is that your tests on inaccurately made miters was implicitly assuming that the inside of the miter was where the gap is...ie the sum of the two supposedly 45 degree interior angles adds up to less than 45 degrees. This places the stress that would be on the inside of the miter on the glue itself. But what if my error is in making those too cuts add up to more than 90 degrees - so that the gap in the wood is on the outside of the joint? In that case, the glue in the gap would be in compression and all of the force would be in trying to tear apart a much better wood-to-wood contact area. I believe that we're in a new era of joint making science because now we have CNC machines that can easily make high precision joints AND which can cut arbitrary curves - we can focus on the joint geometry more carefully - and there is more room for innovation. Some of the "dog-bone" joints that CNC milling machines make can control where the stresses in the material are - and I think that will ultimately matter more than the nature of the glue.

[@DiegoSchmidtfoto]

Thank you!!! A couple of days ago I was wondering about gluing issues and I come across the first video, it has been a pleasure to listen and learn, please continue the series. thanks, again!

[@ThePlendo]

Thanks Patrick, your research is impeccable and enlightening. It has certainly changed my perception of the performance of glue joints.

[@drmkiwi]

More interesting stuff. Another common cause for failure of a mitre joint is the joint being starved of glue. We were taught to apply the glue twice. I don't recall the recommended time frames, but when I'm making a carcase with mitred joints, (and I'm not being lazy) by the time I have applied glue to the 8 ends, it is then a good time to start applying the glue again. In terms of reinforcing the joint, it was surprising to me how much aadditional strength was achieved with a small piece of veneer in a saw kerf cut across the joint. I think Jeremy Broun makes a good point about strengthening joints. Thanks for the video. Cheers, David.

[@whomadethatsaltysoup]

Very interesting video! I, too, like the idea that the skill of the woodworker might be at the root of the weak mitre glue joints. Thanks for sharing!

[@randizzle7659]

I love how much care and detail you do for your videos. So far I've gotten to prove my brotherbwrong on glue joints end to end, but this was eye opening! Excellent job!!!!!

[@morphine4u]

The force direction you use in your test, isn't usually the ones I have a problem with... The forces in a pictureframe or box are also at an right angle to the grain. The example where you break the miter with your hands, are example of this force. Could be interesting to see a test of this... Great video, it's good to shake things up, especially myths. We think we know the truth, but really it's often just cultural handdown 😂

[@jebeda]

I think this is even more of an issue when simulating a "bad" miter joint. I think the plastic spacer slivers should be perpendicular to the applied forces, like they would be if they were in a mitered joint of a picture frame. Thus, in the test rig, you would use a long 3" sliver along the top of the joint horizontally, rather than the short vertical sliver that was used in the test rig. Alternatively one could rotate the samples in the test rig by 90 degrees and test the breaking strength in the other direction, but then you do not get to compare results to the previous tests of end-grain and side-grain as the testing geometry is fifferent.

[@VincentMo]

That's the problem with this test. It tests the strength of a miter joints in a direction that's irrelevant to the forces that most miter joints actually endure. The hand broken test at the beginning is actually more representative to the forces that people worry about on miters than the ones he carefully measures.

[@sparrow_made80]

One of the really great reinforcements of this video for me is in allowing your glue joints time to cure properly. I've had too many mitre failures that are almost definitely a result of impatience. I've recently switched from trying to glue full frames in one hit to clamping up one corner at a time and the difference in success is immeasurable. Also gives you an opportunity to finesse that final joint if necessary. I spline a lot of my mitres, mainly for visual interest, but a practical test of their strength would be a great watch!

[@ABaumstumpf]

"I've recently switched from trying to glue full frames in one hit to clamping up one corner at a time and the difference in success is immeasurable." Gluing them all at once is easy - just do NOT use normal clamps but those straps that you can use all-around. Also, if you need stronger joints than maybe use a spline or a lapjoint (or if you don't like the looks - lapped miter joint) - they make the joint several times stronger.... well, basically anything is stronger than a miter-joint to begin with.

[@EMTDawg]

The parts of the "myth" most people cannot grasp is simple torque... E2E and Miter joints tend to have a glue line significantly shorter than the total length of the project, while side grain glue lines are almost always longer than the width. So the force at the joint is multiplied as you get further from the glueline...This is why Patrick emphasizes not to think about the actual numbers of the break strength...

[@seephor]

Right. the old saying of it's not what you have but how you use it is the key. Wood is sold to us where the grain is parallel to the long side of a board. This is because trees grow much taller than they do wider. So we don't experience wood where the end grain that is longer than the side grain so all of our glueups that involve end grain are short and because of this, we have assumed the reason of a weak end grain joint was due to the end grain but in fact it's due to the small glue surface as well as how the joint is used in a project.

[@ABaumstumpf]

@@seephor "we have assumed the reason of a weak end grain joint was due to the end grain but in fact" No, nobody has assumed that. The endgrain joints simply are weaker. As even his videos show (despite his claims otherwise). He sadly doesn't compare the endgrain-joint to a single longgrain joint in the same application, he always changed the scenario to make the lignin fait instead of using any of the longgrain-joints we know are as strong as the wood it self.

[@miles11we]

@@ABaumstumpf If a joint fails at the glue line or in the wood, that is still the ultimate strength of the joint.

[@subpilot1000]

Your glue studies totally rock! It just occurred to me that the reward incentive created by You Tube advertising will naturally lead to more and more of these wonderful, previously unexplored theories being tested and published. It has already greatly increased the availability of previously closely guarded trade secrets for the rapid advancement of "common tech". Thank you Patrick... I've subscribed! I am appreciating and rewarding all content creators who rock like you by liking and subscribing! The future of education looks brighter by the day.

[@weberwoodshop]

These videos are incredible. Thank you, thank you, thank you. I would love to see you bookend this series with a video that focuses more on actual geometry, aka practical application of joinery. It would me more of a “why certain joints are used in certain circumstances” than a scientific apples to apples comparison, but it seems to be one of the least understood aspects of woodworking.

[@joelwinter4956]

If I recall, it was common wisdom that end grain "starved" the joint of glue. That is, the "straws" of the wood would "suck up" the glue much more readily in an end-grain joint--hence the idea of pre-sealing the end grain with glue, first. Maybe that's really the myth that we're dispelling here, and pre-sealing is clearly not necessary. (But maybe it's something to test, Mr. Sullivan!)

[@ABaumstumpf]

Sadly he does not actually show any longgrain-joints in the same situation cause then it would become apparent why they are used (hint: cause they are as strong as the wood itself). instead he chooses to not compare actual longgrain joints.

[@stefanfyhn4668]

@@ABaumstumpf he already did that in part one. This one is for miters

[@stefanfyhn4668]

@@ABaumstumpf in fact he proved in the first that endgrain joints are much stronger because the bond is more glue than wood fibers or lignin (or whatever the substance is called). This video was to see if the same was true for miters, which are almost endgrain joints

[@ABaumstumpf]

@@stefanfyhn4668 "in fact he proved in the first that endgrain joints are much stronger" Nah, exactly the opposite- he has shown that of the weakest joints that all should be avoided - buttjoints - the endgrain one is the least bad, but he has also shown that it is waaaaayyyy weaker than the wood. "This video was to see if the same was true for miters, which are almost endgrain joints" Yes, and again he didn't show anything but the weakest joints that everybody (contrary to his claim) advices against using cause they are just that weak. Even funnier - this would be the perfect candidate for a splined miter (way stronger than a normal miter) or a lap-joint (an actual longgrain joint)

[@gordroberts53]

Very happy to see a follow up to the first video. Exceptionally well thought out and presented. I’m a little wood-wiser as a consequence, thank you!

[@burstofsanity]

Another excellent video! I would say that most miter joints that people are concerned about being weak are for boxes and picture frames where the relative size of the ratio of glue area to leverage length is much smaller than say a table or cabinet top with a mitered surface done for decorative purposes. Also, most of the failures in the real world are at a different orientation than the lab tests here. I can't think of a simple way to adjust the test to be more in line with the hand demonstration.

[@dennisd5776]

You are doing a great public service in producing this series. Supplanter ask why so many people with a math or engineering background are in woodworking. I believe that both background are creative or build something oriented. I also feel that this is why we can have an intelligent discussion on this subject, as we understand the scientific method being used and then the resulting conclusion's. I am looking foreword to the next video.

[@testbenchdude]

I don't have much to add other than my observation that your charts are astoundingly clear and easy to understand. I think my geometrics professor would really appreciate seeing this (the class was heavily focused on best practices for displaying data in a meaningful manner). Thank you once again for doing these tests and letting us see your results. Looking forward to more!

[@jeffsmith5084]

Just as good as the first video. Carefully thought out and clearly presented. This series is going to be bookmarked.

[@RabbitInAHumanWoild]

Thanks very much for this trilogy that I hope you will expand. You have influenced my decisions with respect to joinery. I first started to question the common wisdom about the weakness of mitre joints a few years ago when I was making picture frames. I messed up on the size of one and thought that I would be able to pull it apart and dispose of it very easily. Pulling on the sides in an attempt to break it apart was much harder than I expected and left me wondering what was going on although I never followed up my observation. Thanks to you we now have quantitative and qualitative information about glued joints and no longer have to follow rules of thumb that are probably best described as mythology.

[@Guruc13]

Thank you so much for this. You're changing woodworking forever

[@GarageWoodworks]

Modulate the moisture content (MC) of glued samples up and down and quantitate w/ a moisture meter? Examine how repetitive cycles (MC) influence miter strength? Loving this series!

[@cymeriandesigns]

Thanks for your efforts in making these videos. If you run out of other things to test, I wouldn't mind knowing where baltic birch plywood joints fall on the strength scale compared to solid wood.

[@Idefix70]

This would be an interesting test as plywood has alternating grain directions!

[@robandsharonseddon-smith5216]

@@Idefix70 And the laminations will likely separate too.

[@dwightl5863]

@@robandsharonseddon-smith5216 I'm a real novice here but are you saying the laminations, the wood itself, would separate or the glue line holding the laminations together? The glue is stronger but now factor in how impregnated the wood is with the glue being those laminations are quite thin.

[@robandsharonseddon-smith5216]

@@dwightl5863 When plywood fails, it usually does so by delamination, failing along the glue line between the sheets first. It will be interesting to see what effect if any that has on the strength of joints made this way. I don't think it will be practically significant - this series is not going to change how people do woodwork but it is very interesting.

[@sawdustadikt979]

As a renovation/trim carpenter in the field for 30 years. I’ve done some experiments on miters as well. In my experience it seems that the difficult conditions that southern New England provide really test your methods. I prefer to only build hard wood decks. Last deck I did was all ipe. I splined all my withers. I cleaned all the glue surfaces with acetone and used a total boat epoxy that is made for oily woods. I picture framed the deck, the stair treads, the hand rails and the built in bench. I tried bedding the picture framed deck in pl and used a lot of screws to immobilize the joint. I used table top hardware to fasten the handrail joint to let them move and the stair treads I used over sized holes with plugged fasteners to let them move as well. The desk has never been oiled. Here are 4 years later, the only miters that cracked were the deck picture frame that was glued and screwed to the point of excess. Everything that was allowed to move is still as tight as the day I did them. That made my mind up for me. I love the idea of really getting to learn more about woodworking, it really makes me a better carpenter. I don’t mind getting out some hand tools to do something, less dust and it’s almost as fast as power tools.

[@davidneale-lorello2954]

Absolutely loving these videos! I can only guess the amount of work that goes into the research and video production. Thank you for the rich and practical insight you offer the community!

[@davcar872]

Very informative video. There's another KZbinr who's a commercial finish carpenter that has done this type of video using CA glue. In every test, the joint never failed. It always failed elsewhere on the piece.

[@FrankC76]

I’m going to enjoy the wave of response videos to this one. I, for one, will keep using splines on mitered frames and boxes because they look darn cool.

[@billyjack8906]

I wouldn't suggest to ditch them, myself. As long as the splines are long-grain against the 45°, they will significantly strengthen the joint, regardless of how "perfect" the miters are cut and assembled. The long wood fiber is still much stronger than even the strength of the glue. Due to the small surface area of bonding, on a typical mitered corner, most still benefit greatly from that addition.

[@atscscience]

He alluded to preliminary results of tests with splines... But yeah. I hope he does continue the series to include interlocking traditional joints. It kinda sounds like he plans on doing much more Makes sense given where the series has started.

[@atscscience]

I'll still do traditional joinery, though. So pretty.

[@ABaumstumpf]

Well, you are doing the right thing cause the miter-joint is basically the weakest, followed by an endgrain-to-endgrain buttjoint. Using a spline will seriously increase the strength, or you could also use a half-lapped miter (if you do not like the look of a simple lapjoint) - both waaay stronger than a miter.

[@CarlosRodrigues-rl8px]

Thank you so much Patrick for spending the time and sharing the conclusions.

[@laroseauxboisrosewoodwoodw8585]

Oh boy finaly someone is telling the truth !!! 32 years ago , at my woodworking school , i’ve lerned EXACLY what you’re showing with your tests 👏👍🏼 When i was saying that the end grain was strong , i was laughed at etc … Now i’ll point ouf your video and finaly have something to back me up 😜👏🇨🇦❤️

[@MichaelOlsen-Engineer]

Yet another exceptional video using the scientific method to derive the facts about something we thought we knew! Well done Mr Sullivan,. Well done indeed! I presume from your methodology, maths, and commentary that you are an engineer, and most likely of either the structural or mechanical disciplines. It never ceases to impress me as to the large number of engineers who seriously enjoy wood working and who are willing to share so much of their professional talents in an essentially pro bono fashion. I am wondering if you have looked into the history of glues used in woodworking and considered how their properties would have effected our understanding of glue joints? For example, hide and fish glues slowly crystalize. These glues are referred to as cured based upon tactile observations and perceived adhesion. However, close examination of antiques built with such glues has shown that they do in fact crystalize and it is the hardness of these crystals and the strength of their fracture planes that provide the strength long after the plasticity of adhesion has gone. Such joints, in say old Windsor chairs, are often subject to multiple and varying stresses which help to "grind" out the crystalized glue. Restorers often use the slow crystallization process and heat sensitivity of these glues to their advantage. The joint in question is carefully heated to re-liquefy the glue, and if needed, more glue can be added to replace any lost over time. The result is a restored joint that is virtually indistinguishable from the original. PVA glues on the other hand maintain a significantly more elastic plastic quality, They are able to "move" with the joint longer than older glue types. Where as CA (cyanoacrylate) glues crystalized extremely rapidly. The have exceptionally strong adhesion and cohesion, but are highly susceptible to break down from vibration and thermal change due to their more brittle crystal lattice. Equally important is that the basic formula for all glues have changed considerably over time, especially the last 50 years or so. This means that what was true for craftspeople of the past, "today's teacher", may no longer hold true do to those changes. Unfortunately, the only source of information most have had available is what they were once taught and what they have experienced. Therefore, the observations of the past may well have been impacted by the conditions of that time, which would include different glue formulations. Since all teaching is based upon history, something readily show as a myth today, is often based upon relatively sound advice, and multiple factors, for its time of origin. I mention this as yet another set of conditions that explain why only relative, not absolute, comparisons make sense in the cases of your experiments. Personally, as an engineer myself, I can find not fault in the design of your comparative experiments. In fact, I could easily see where they could be developed into either an M.S. thesis or Ph.D. dissertation with the adhesives, lumber, and carpentry industries being very interested in providing grants. Thanks again for producing some incredible videos. Here is wishing you and yours the very best.

[@yosoyluismendez]

Great content! One of the most impressive and useful resources I've found on the internet about this subject. Thanks!

[@ronlokaisingh9344]

Wow, love your channel Patrick! You speak to my love of woodworking and my engineering background. Keep up your great experiments and videos. I very much appreciate your excellent work.

[@westernbrumby]

I look forward to your future videos! I will come back periodically to check.

[@bartojoh]

Thanks again for taking so much time and care to make another complex subject seem deceptively simple and easy to understand. I'm sure your videos will become legendary in the realm of youtube/woodworking presentations.

[@bobfraser5757]

Fabulous work! I fully endorse and appreciate your effort. It has erased some frustrating questions that bothered me for some time. Please keep the videos coming .

[@doubleedgedsword6352]

I used to just think you were mistaken, owing to a poor scientific theory. But then you called yourself an engineer, and now I at least just think you're a contrarian doing it for the subs. Glad to be able to sign off. Looking forward to your next video, 'Concrete Is Weaker Than Wood!!'

[@ICUNA22]

Excellent!! I like the fact that you stress woodworking skill; i.e., accuracy of cuts, as a major factor in the strength of the joints. This video, along with the previous one, was a real eye-opener for me. Thanks!

[@lonpetchenik3742]

Echo all below.. your videos are making me question about how I think about (based on my readings) and discuss the 'dogma' of glue joint strength.

[@jeffj2495]

Wow, actual viable content on YT. Who would have predicted that? Another great analysis, great testing, and a very clear vid showing the results. THANK YOU.

[@oldfreddyfrenchfry1]

Fantastic job - I love these! My bet about the gapped results is that minor gaps can still allow the glue to connect the two faces of wood, kind of like mortar.

[@pengel200]

I've enjoyed both your test videos and the commentary from the likes of Marc Spagnuolo and Stumpy Nubs that have resulted. The testing is very well done and driving a lot of thoughtful science. On this video I caught a glimpse of your watch. You have great taste in time pieces, too!

[@robertanderson2223]

Once again, love you analysis and approach. Thank you for the insight! I am curious to see the effect of splines on these miters.

[@baltasarpuig8754]

Thank you very much for adding a little common sense and empirical treatment to a very common subject riddle with misconceptions and carry overs from forgotten sources. Glue is effective in endgrain. The problem seems to lay in the definition of how how effective this is. Long grain glueups are as strong or stronger as the wood itself and that is the expectation. End grain glueups are many times weaker than uninterrupted long fivers. End grain joints are the wrong choice if you have the expectation of continuous strength over the length of a structural member. If you are doing marquetry, glue away!

[@pjhalchemy]

Enjoying this series immensely Patrick, Thank You. Personally, I like that it has stirred the community into thinking about these things in more depth and bringing on Conversation. Lots of valid points below, particularly the typically closed loop, but your apples to apples approach is quite valid and thorough, imho. Joint strength is important for longevity and applying atmospheric changes is part of that, but with all multitude of joints I think of Japanese temples in the elements lasting 500+ years with simple and ingenious locking mechanisms and no or little glue or pins to be the ultimate in strength and longevity and wood choices...especially in heavy structural elements. Looking forward to more of your fine work on this subject! Hat Tip, ~PJ

[@babylonfive]

Isn't it true that nearly all miter applications are where the surface area of the miter joint is significantly smaller than the largest dimension of the final item? This returns to the idea that leverage allows assembly-wide stresses to overcome the small miter joints. Examples include boxes, picture frames, mouldings, etc.

[@Michael-Makes-Stuff]

I love this series, Patrick! So interesting and educational. Also creating some great discussions in the woodworking community. Keep it up! 👍🏼

[@atscscience]

Again, excellent. Can't wait for the next one. Still wanting to see how wood movement stresses the glue, though... Make an anti-kiln that controls humidity and make the relative humidity oscillate with a period of, say, 7 days, for a month or two or twelve? This might be a way to simulate many annual cycles in humidity in a reasonable period of time. - From an atmospheric scientist and woodworker

[@atscscience]

Actually, I could provide typical quantitative annual, synoptic (i.e. variations due to weather systems), and diurnal variations in humidity quantities of your choice for selected regions if interested.

[@benwiley3962]

This makes a lot of sense and is extremely helpful to see "how accurate" is "accurate enough"...thanks for sharing! Could you possibly simulate seasonal wood movement by placing the glued test pieces in a high humidity environment for 1 week them moving them back to a low humidity environment for another week to guage the difference in strength? obviously doesn't simulate years or decades perfectly, but might be a step towards proving/disproving the seasonal movement counter argument.

[@bobby-c7731]

Mr Sullivan. You’ve earned a subscription from me with your insightful and thorough analyses. And, I hereby name you Chief Pot Stirrer of the Woodworking Community! I’m looking forward to the 10 or so responses that will be gracing my feed.

[@pawpawstew]

I can appreciate the testing from the point of "Here's what happens if your joints aren't tight.". It seems, though, that the root of the issue is having the correct tools to cut miters, how to ensure that the tool cuts accurately and how to correct any inaccuracies, and finally how to use said tools effectively to get the results desired.

[@MartinPineda98765abcd]

Thank you very very much for this series sir. This series is important not just for woodworkers but to so many other disciplines!

[@mcgeorgerl]

For a long time I have purchased picture/artwork frames at auctions since: A. Commercially available wooden ones are disappointing and B. Commercially available ones do not fit with our decor. So I buy frames likely 50 or more years old for reuse. Most of these frames are (presumably) hide glued, cross nailed and then the nail holes are filled. The number of miters that have failed glue joints are probably 75% or higher. The only things holding a large number together are the cross nails. So I hack-saw through the nails, drive them out then recut the miters down to fit the artwork/print/photo at hand. I use an accurate, dedicated miter cutting fixture. Depending upon the size, I may just glue with Titebond original and cross nail again. These should last the rest of my life. But with larger, heavier works... I will always further reinforce the miter. I'm working on mounting a rare 24" x 36" county map so with frame, UV glass and plywood backer (paper/foam archival mounting supplies are not a significant weight) so I'm dealing with a fair amount of weight hanging from stainless wire mounted at two points. Very interesting video. I'll keep an eye out for any regarding splined miters. Love the methodology.

[@ilikewaffles3689]

I love the fact that everyone was quick to change their ways in the face of evidence (your last video). Really amazing stuff.

[@ABaumstumpf]

"I love the fact that everyone was quick to change their ways in the face of evidence (your last video)." But all that video has shown is that glue is stronger than lignin, but he intentionally did not use any longgrain-joints like lap-joints (which would also have been perfect here). A good joint is as strong as solid wood and he shows that endgrain-joints do not come even close ........

[@ilikewaffles3689]

@@ABaumstumpf I'm pretty sure he said he was only going to test end grain vs lignin/edge grain. Lap joint is not testing only lignin.

[@ABaumstumpf]

@@ilikewaffles3689 In the first video he said he would be testing endgrain vs longgrain joints and then even said that endgrain-glueup would be the strongest joint. And here at 6:36 he even claims that butt-joints would be common and nobody warns against them - which is the exact opposite where everybody will tell you to avoid them for anything but decoration as they are extremely weak and the glue is prone to cracking.

[@Isaiiahii]

I'd like to see how strong a miter is with a #20 biscuit added. I bet it will be quite a bit stronger even though they claim biscuits do not add any strength and are only meant to line the wood up.

[@L__Y]

I never understood the claims that biscuits don't add much strength. They're like plywood splines so one would always have some number of layer with favorable grain direction regardless of shear load direction. I guess in tension it becomes a function of the additional gluing surface added.

[@monteglover4133]

Miter and end grain joints were the accepted exception to the “rule”

[@egilsandnes9637]

Yeah, I guess when glueing to pieces of MDF together (edge to edge) the biscuit wont add much strength. In that case the glue joint is much stronger than the boards.

[@DavidLauritz]

I love your humble and scientific approach to this! We have to many people who are to secure in their opinions. You only seem sure that it all depends... =)

[@nonexman]

Patrick, again a brilliant video, thanks. When I watched your first miter joint break under the machine, I immediately recognized the issue: failure finally went to the weakest part of the joint. And in this case, it's the lignin. But to clearly focus my comment, I have to revert back to my work of over 40 years (until I retired 6 years ago), I was a scientific glassblower who also did a lot of research and study in the field of the physics and properties of glass. To break glass, you need two things: strain and a flaw (in the glass). Glass that is flawless is incredibly strong (stronger than steel). That's easy to understand as there's no place for breakage to initiate. However, what's surprising is that the shape of the flaw can make a big difference in the vulnerability of the glass. If you think of flaws in the glass surface shaped like a "U," a "V," or an "|," each of those flaws has a different radius at the bottom. The strength of the flaw is determined by its "K" factor where K= 1+ 2(A/B) where A is the depth of the flaw and B is the radius at the flaw's root. With that in mind, you can see how a "U" has a very low K while an "|" has a very high K. So, what does all this mean in practicality? When people are trying to cut a sheet of glass (for the first time), they'll often run the cutting wheel up and down the glass several times to make sure they have a wonderful, deep scratch. (AKA, a "U" cut/scratch.) However, at the same time, they've created little micro-scratches that go off to the side. When they start their cut, their initial strain at the start of the scratch is focused on the deep "U," but as the crack wanders up the scratch, as soon as it hits a micro-scratch, with its greater K factor, the crack eagerly wanders off to the side, away from the intended scratch, to the novice cutter's consternation. So in your case, the initial failure was the glue UNTIL the crack came to a weaker area, the lignin. Now I'm not saying that the failure (breakage) of glass, glue, and wood is the same. However, the dynamics at play here I believe are: failure will occur at the weakest link which can vary depending on where the strain is at any given moment. But again, thank you for a wonderful presentation.

[@thedistractedmaker]

Great video Patrick. Often we get hung up wanting our projects to last "forever" and knowing glue will eventually fail doesn't sit well- so we over-engineer as much as possible all while avoiding dreaded nails or screws...

[@slowdaze]

I think you have shown for your self why we say endgrain joints are considered weak. 1) in your first video shows that on endgrain glue ups the glue fails, where as on side grain the lignum fails. Thus on side grain the glue joint appears to be stronger than wood and endgrain weaker than wood. 2) While 3” glue ups would be common on endgrain, sidegrain would tend to be much much longer. I would think this would distribute the load over a greater area. 3) Geometry maters in the real world. Miter/endgrain would more likely be a small glue joint with a much longer lever overhang on either side of it. Side to side would more likely tend to be a long glue joint with much less of a lever working against it.

[@ABaumstumpf]

"3) Geometry maters in the real world." Yep - and that is why the miter-joint and end-to-end glueups are the weakest joints (contrary to his claims), as a good longgrain-joint maximises the overlap of the fibres while the glue replaces the task of lignin, making the overall joint stronger than solid wood.

[@hydorah]

Looking forward to the testing of maybe aligned and perhaps diagonally cross grain splines If you could work out a way to reliably produced grain aligned splines for testing end grain joins and diagonally crossing splines for mitres that would be amazing. This research series is already amazing and you , Sir, are doing incredible work! Thanks. I also love the way you listen and incorporate suggestions into you commentary and future test plans. I also love the way people comment on this series. Rare on KZbin you can learn additional things in the comments!

[@nirgosh]

This is fantastic, thank you! I'd think it's equally important to measure the strength of the joint in the 'other's dimension, that is, applying the force on the short sides rather than the faces. The reason is that any force that is applied on a miter frame would act on the corners, forcing them away from 90° in both +/-.

[@howycwap]

People say miters are weak because they are. No one talks about side grain glue ups because no one glues up a sidegrain board as short as a typical miter joint. Typically it is always said about picture frames where they are usually at the end of a long torque arm

[@anthrazit6262]

Great video, thanks for the insights! I believe another reason people tend to underestimate the strength of miter joints is simply its use case. Using side-to-side glueups in a tabletop implies very long glue joints that (obviously) cover a lot more surface area than a mitered picture frame for example. Seldomly would someone build a tabletop where all the panels are cut at 45 degrees to the grain direction of said panels. Now we know that if they did, that tabletop would be inherently stronger at the joints. It would just take some veeery wide slabs :D Keep it up!

[@ABaumstumpf]

"believe another reason people tend to underestimate the strength of miter joints" Nah, it just is weaker. Build 5 90° joints: butt-joint and miter - by far the weakest splined miter - already stronger as you introduce more longgrain lap-joint - way stronger box-joint - the strongest - here the wood will fail before the joint.

[@meraydin1]

I hope you'll be around to prove anything you wish. You are #1.

[@GreenOne01]

Wow! This is a really good assessment. I appreciate your thoughtfulness & testing methodology, as well as your interpretation & presentation of the results! 😁 Thanks for sharing!

[@davidburris120]

I’d very much like to see a video like this with splines! Thanks for your work!

[@briantaylor9266]

I'm really enjoying these videos, and the responses from the big names in the YT woodworking community. As an engineer, I especially like your rigorous approach to the testing. We have always been told that it really doesn't matter what glue you use because it's all stronger than the wood. Your demonstration that this is not true for end grain and miter joints begs the question what glue is best for these. PVA, epoxy, polyurethane, hide?

[@ABaumstumpf]

"Your demonstration that this is not true for end grain and miter joints begs the question what glue is best for these." It shows that endgrain joints are weaker cause actual longrain-JOINTS are, as you said, as strong as solid wood, he just chooses not to show any longgrain joints - why ever that might be.

[@spruce_goose5169]

@@ABaumstumpf You might want to watch his previous 'glue myths' video.

[@ABaumstumpf]

@@spruce_goose5169 "You might want to watch his previous 'glue myths' video." I did - and what it showed was that buttjoints are by far the weakest and the endgrain-joint was just the least-bad out of the worst, failing at not even 50% the strength of the wood, were as something like a finger-joint is as strong as the wood. Really, just watch the video and then ask your self why he is only showing butt-joints - the weakest known joints.

[@spruce_goose5169]

@@ABaumstumpf 'Strength of wood' is too nebulous without specifying the axis. Your response is nested under Brian Taylors', which states that the strength of the end grain glue joint depends on the glue. A side grain joint fails at the lignin bonds, so there is no benefit to using glue any stronger than standard wood glue. But since end grain joints fail at the glue, a stronger glue may be of benefit. There is no argument here against the reality that an end grain glue joint is weaker than in-tact long grain fibers. The takeaway from the first video is that end grain glue joints are stronger than lignin bonds. That takeaway does not provide logic against joinery, of which I agree is still needed. It simply says that end grain glue will hold longer than lignin when loaded across grain. Do with that what you will.

[@ABaumstumpf]

@@spruce_goose5169 "'Strength of wood' is too nebulous without specifying the axis." ..... are you kidding me? Or are you really that dense? "But since end grain joints fail at the glue, a stronger glue may be of benefit." Yes, and using a better joint than and endgrain butt joint would be even better - so much better even that it would be no longer the glue failing but the fibres and lignin. "The takeaway from the first video is that end grain glue joints are stronger than lignin bonds." Yes, and not as the video claims that endgrain-joints are string.

[@drooplug]

Your work is certainly interesting and does shine a light on this topic from a different angle. I would guess that the idea that end grain joints are weak has its roots in the era of hyde glue. I've been a professional woodworker for over 20 years. It's difficult for me to think of many clear instances of failed end grain joints. Mostly because I have avoided them. When I think about the many items I have disassembled, I can say that my experience is that the end grain joints fail more easier. Yes, there are times when end grain glued to face grains holds well, but I haven't found it reliable. The two examples I can think of are coped joints of a cabinet door and glued edges of plywood in a box. In the door example, the joint at the end grain portion is always much easier to break on the glue line than the face to face joint at the tongue. In the plywood example, when break apart a piece of plywood with its edge glued to the face of another piece, the end grain of the edge almost always remains intact and the face grain breaks out. On the topic of miter joints, I was taught that they are in-between end grain joints and face grain joints. I. The instance of a picture frame, the strength of it is 100% derived from the joint at the miters. Just glued miter isn't strong enough to resist those forces. However, mitered moulding attached to a piece of furniture rely on more than just the glued miter joint for strength. It is attached along it's length elsewhere. I will also add that clamping is very important for strength. Miters are notorious for being difficult to clamp effectively.

[@theeddorian]

The one question I would really like to see addressed is when force is directed through width dimension rather than thickness. Loading on a picture frame for example tends to be perpendicular to your experimental setup. Also, glue type may have a distinct effect on whether wood or glue fails first. I have the pieces of a splined, oak frame made by my grandfather during WW I. The frame essentially disintegrated over time as the hide(?) glue he used aged and became brittle.

[@ronhochhalter3491]

Great videos, when you get all the "other guys" making videos in response to your videos, it makes for healthy conversation. Can't wait to see the next installment.

[@SnootchieBootchies27]

I really like your logical, actually scientific approach. I do think that seasonal movement is a factor with miters on wide boards, say 4-5 inches or wider, depending on species. I think the amount on movement on small pieces like the ones in your clock is not going to be enough to compromise the joint. My rule of thumb is that if I have to start thinking about putting two biscuits into a miter (because of width), it's probably not going to hold up in the long run anyways.

[@Richard_TheCreator]

Real-world results, that everyone can generally rely on. Thank you for doing this.

[@ABaumstumpf]

Sadly no comparisons to actual longgrain joints - yet again.

[@colinellicott9737]

Great series. A standard practice in reliability is to cook the item at a carefully chosen elevated temperature to simulate aging (Arrhenius), coupled with humidity cycling this could accelerate the testing instead of waiting 50 years (HALT & HAST). Again - thx for all the practical insight 👍

[@hardnox6655]

Well done Patrick. Thank you for sharing. Great information.

[@libertarian1637]

I like the scientific testing and the explanation in this video of the idea of torque. When you test the miter or an end-to-end glues joint, or others demonstrate it, it’s always with a small joint and long legs of wood that seemingly don’t need much force exerted to brake the joint; this however is simply torque with a small force at the end of a long lever translating or equaling a large force at the joint. I’d love to see a comparable glue up of side-to-side with end-to-end where each piece was say 12 inches long over a 3 inch joint. I would think this may put some of the arguments and negativity to rest as this is the strength of science that people just can’t seem to wrap their heads around as side-to-side glue ups are so standard and no one ever places them under the same force/stress that end-to-end glues face. I tried this myself with some old growth oak and hard pine and could definitely tell that the end-to-end was at least 3 times stronger than the side-to-side; the side-to-side broke without much effort over the 2 feet separating the grain while the end-to-end only broke when the glue gave out.

[@NeilIsaac]

Cool to see how the joints fail under a worst case load. I would be interested if you could compare the failure point of end grain vs side grain glue. I think you could do an apples to apples test by cutting a rabbet in the end grain of the two blocks then gluing either the end grain surfaces, the side grain surfaces, or both. Since it would be aligned the same as the end grain joint, the glue should fail before the wood fibre so we can find out whether end grain or side grain glue is bonded stronger. Cheers.

[@onebackzach]

I'm sure that many people have already recommended it, but moving glued pieces between a damp environment, such as a unheated basement, to a dry environment multiple times over the course of a few months might sufficiently model seasonal expansion and contraction for the purposes of the video. It might also be interesting to leave glued pieces out in the elements and document how rain, snow, etc. weakens the joints.

[@billm2158]

Great video and insight again Patrick. Thank you.

[@PaulaBean]

I love the scientific approach.

[@travel734]

A good video that will certainly give rise to a lot of comment. The series, thus far, is also a reminder that adhesives are not joinery. We need to remember to consider the loads (static and dynamic) and then design accordingly. Mr. Sullivan hinted that he may look at splines in the future, hopefully that will only be the start.

[@samz8023]

Fantastic as always. I do think that the moment of force applied during the test should have been about the axis perpendicular to the face of the boards

[@ingricardomaldonadoveronic3500]

Congratulacions Come back!!! Bravo !!! Mr Patrick

[@kren1260]

It's absurd this kind of experimental look and in depth discussion in carpentery techniques didn't already exist. But I'm glad it now does! I will be following religiously so I have the leg up on all the other carpenters I know

[@raphaelklaussen1951]

Of course they exist, people just dont bother to read them. The US Forestry has published manuals and papers with that info for many years.