@@DanDDirges hell yes! I didn’t even know that I wanted Marc’s take after seeing Patrick’s video, but here I am... the challenge, however, is adding something to the topic rather than just doing another ‘version’. Good luck!
Going on 50 yrs ago when I was a teen learning basic woodworking, I discovered that it actually DID help to glue things like the end grain shoulder of a tenon, or the end grain of miters, etc. More recently it’s pocket hole face frame joinery and the like.NOT because it necessarily makes anything “stronger” overall, but because it DOES keep all the joinery looking better throughout expansion/contraction cycles and typical daily use, to the point of having far fewer cosmetic issues like hairline cracks in finishes or dirt in the air discoloring the unglued edges over time. Your glue is already out so It takes a couple extra seconds… why NOT do it?
I watched Patrick's video a couple of days ago and until you mentioned it I did not think that anyone would think that you could do away with mortise and tendons. I was wrong again. Great points as usual Marc.
Sullivans Video ist very good. So is your explanation. Thanks. As an furniture restorer i see every day what joint fails. And it is the endgrain joint and the wiggly/loose joint. Never rely on glue. The Elder secured every joint with some sort of mechanical interlock like pegs or dovetails for a reason.
Great video and a shining example of respectfully adding to someone’s content. I really appreciate that you took the time to clarify at the beginning and end of the video that you agreed with him. You are saying that the myth is a fallacy, but it is still of little practical use. I also find it somewhat annoying that you have to make it that clear so people don’t twist your words.
No need to twist his words--he downplays Patrick Sullivan's remarkable result. He said Patrick was right BUT who cares. Instead he should have said Wow Patrick has flipped a standard claim of 100 years AND here is some clarification for those few people who may have misunderstood it's impact to joinery.
Man I'm glad you came out with this video before trash day! Was able to get my DF500 and 700XL out before pickup! Seriously though I enjoyed his well made video (and his others) and was interesting to see the breakages along grain lines but it changed nothing about how I would approach any joinery.
Loved this video. I was thinking about wood movement the whole time end grain to edge grain was discussed. It would be interesting to see those tests after the wood has gone through a year of humidity changes.
Excellent couple of videos by both TWW and Patrick. I came to UA-cam trying to figure out a joint I’m trying to make so the timing of this is excellent. If I use my Domino to make a joint where one end of the Domino sits in the mortise such that the long grain of the domino mates to the end grain on mating piece. The joint I’m trying to describe is like joining the sides to the back of a box. Would that work? Patricks video seems to imply that it would? I really don’t want to use screws or pocket holes, but there is no other way to orient the domino for this particular joint.
I was surprised by Mr Sullivan's conclusion when I watched his video. Then, you bring your knowledge and, applied to real world woodworking, everything become obvious: mechanical strength! Thanks Marc, it's your turn to make me feel more intelligent 😂!
Great follow up video. I agree everything Patrick said was correct but it doesn't really reflect what we generally do in woodworking - although perhaps it indicates we should change some of the things we do. The one thing that I'd like to have seen mentioned here though is the way the joint fails. With an end grain to end grain join the failure is sudden and catastrophic. Long grain failures, in my experience, usually give you a bit of warning before they let go and mechanical joints tend to give quite a lot of warning.
Thank you! This was my precise reaction to the video and some of its responses - albeit better said and available to a much wider audience. His findings completely (and somewhat shockingly) bust the myth. What they don't do is invalidate the hard-won lessons of hundreds of years of furniture makers vis-a-vis functional joinery. One reason the myth has stood for so long is that there is almost never any good reason to glue end grain to end grain - if you want to stretch a board you'll typically use some form of scarf joint which increases the glue surface and hence the strength.
The myth does make sense in the context of, if you're gluing side-grain to side-grain, you end up with a joint that is stronger than the rest of the board, and if you're gluing end-grain to end-grain, you end up with a join that is 1/4 to 1/8 as strong as the rest of the board. I don't think most people realize how relatively weak wood is if you stress it with the grain.
This is a point that only us structural Engineers probably appreciate---- Stress or pressure is not applied to a joint----- A. Force is applied to a joint----- Stress and pressure are the result of that force., and are highly dependent on the area that the force is spread over.
In joining end grain to side grain, apply a thin coat of Titebond to the end grain surface and let it dry before applying the glue to the two boards to be joined. This seals the end grain so it won't suck the glue away and starve the joint.
Completely agree. When using titebond 3 I change it up a little by not allowing the primer/sealer coat to fully dry. The additional dab of glue before assembly can then make a better chemical bond.
Patrick was just busting the end grain myth. The real takeaway is that if you would be OK with a joint using edge grain, there's no reason to think it wouldn't be OK with end grain. All the other things you would normally consider -- grain direction, wood movement, need for hardware, etc. -- still apply. To add a minor thing, though -- the direction of the grain is also important when considering how hardware such as screws are used. Screws into end grain are not as weak as people often assume, but it can still be a bad choice depending on the application.
"The real takeaway is that if you would be OK with a joint using edge grain, there's no reason to think it wouldn't be OK with end grain." This nicely encapsulated the important lessons. If you took a square frame cut out of solid wood and subjected it to load, you'd expect the failure to occur along the short sections of the grain. So yes, glue is stronger than wood, but wood itself isn't strong enough for the job if you're relying on what amounts to "end grain boards". Also absent in must of this discussion is leverage. Sure, 400 lbs sounds like a lot, but that load is trivially reached by the mass of an adult placed out on the end of a lever. The joints of a chair in which someone is leaning backwards is subjected to many multiples of their weight.
I think it was an excellent video by Patrick, however my biggest gripe is that it's not really comparing the strength of a glue joint along the grain versus a glue joint with end grain. It's only really proving that the glue along the grain is stronger than the wood fibers themselves, as the wood is what broke in every test, as opposed in the end grain tests it was the actual glue joint that broke. I think that confuses a lot of people as well, even though that one of the tests do not show a failure of the glue joint, quite the opposite.
No, it did not show that the glue is stronger than the wood fibres! Quite the opposite, in fact. What his tests showed was that the glue is stronger than the LIGNIN holding the fibres together, but not as strong as the fibres. This is why long grain glueing broke the wood (ie the lignin bonds), and end grain glueing broke the glue. It took more force to break the end grain glueing exactly because the glue was stronger than the lignin.
Since the side grain glue joints all failed AWAY from the joint the only conclusion you can draw is that this glue Joint is at least as strong as the cross grain strength of the wood. Because it never reached failure in the glue joint, you cannot say that it is weaker than an end grain glue joint
@@jeffhigh2 No, this is not entirely true. Let me try and explain what/how Patrick Sullivan (PS) measures: Initially, there are two variables: glue strength and wood strength. However, for this purpose, we can break down wood into two sub-components, fibres and lignin: Fibres are cellulose tubes that are the transportation means for water and nutrients in the tree. Lignin is the glue that holds these fibres together. Thus, we can talk about glue strength, fibre strength, and lignin strength: Glue strength is measured as the force it takes to break two pieces of wood along a glue joint (S^g). 'Glue' is here a pva-glue. Fibre strength is measured as the force it takes to break a piece of wood across the fibres (S^f). Lignin strength is measured as the force it takes to break the fibres apart from each other, ie break a piece of wood along the fibres (S^l). From the empirical tests performed by PS (watch the video!), the only general information we can infer, is that glue is stronger than lignin (along grain glue-ups are bound to break away from the glue line): (S^g) > (S^l) End grain to end grain glue ups do not provide any relative information about strength even though they are bound to break at the glue line, since the wood fibres have already been severed: * => (S^f) > (S^g) Thus, end grain glue ups provide glue strength information only. Luckily, PS does not rely on relative information only, but he also performs absolute tests: he measures how much force it takes to break apart two given test pieces. He tests a number of wood species for strength in different directions (Sw^f and Sw^l) as well as glue strength for the given wood species, tested as end grain glue ups (Sw^g). These tests all reveal that: Sw^f > Sw^g > Sw^l For all tested species and specimens, fibres are stronger than glue, and glue is stronger than lignin. The actual force required differs from wood species to wood species, however all tests show the same relative order, namely that end grain glue ups require more force to break than along grain glue ups. This is what is illustrated in the graphs also shown in this video. Remember, these are test tube results, and they do not tell us anything about if or how we should change the way we work, as is the gist of this video, but they DO provide us with a piece of information that is most likely new to most of us. Bravo Patrick!
@@pinkerbot But the test never actually breaks a side grain joint, the material breaks before the glue fails. You have effectively just measured the cross grain bending strength of the material (Which is enhanced at the joint by glue penetration) What the test tells us is that an end grain glue joint can be as strong as the wood cross grain in bending (glued or not)
hello, What i'm missing here is that when you bring a piece of furniture to a client/ buyer you want it to be as strong as possible because when it breaks or get damaged because it's made to weak you have to go back an fix it, you never have time for that and the client is sort of disappointed. So my point is: engineer the piece in a certain way that it's durable. It's already pricy when you make handbuild stuff so expectations are higher. Still it is really good to know all this background information about the strength etc.. really helpfull and gives you a bit of confidence when you make stuff and you think yeah it's durable enough. greetings from Europe
Great job adding context to this topic. I loved you highlighting your concern with what viewers would take away from Patrick’s video, not his video per se. The good YTers realize it’s virtually impossible to have the message a creator intends to convey be the one received by all that watch it. He does mention that a future test will be on mechanical joints. I’m looking forward to that one.
Thank god someone has some common sense. People are acting like this video destroyed centuries of wood working techniques and wood workers are doing it wrong.
Great video Marc. We've been inundated with questions on this lately and like you, I can't really come up with a real-world use for an end grain joint-save one, on the podcast we answered a question about solid-wood chess boards and Patrick's video would make me rethink my answer on that one. -Ben
Another end grain to end grain use... well... kind of the same one, create a pattern. ua-cam.com/video/NUTmBy-63qo/v-deo.html My only real conclusion from Patrick's video was, if my glue up allows the time for it, I will now add glue to end grain in addition to just long grain to long grain contact surfaces.
@@edu_carceller I'm not sure that's necessary. We've all collectively been building furniture using white and yellow glue for decades now, and for the most part if the joinery is sound, the furniture holds together just fine. I'm not convinced adding a little extra glue strength is going to make much difference, when the bulk of the strength comes from the joinery itself.
@@Mikey__R Oh I am sure it is not necessary because tons of woodworkers of all skill levels have been avoiding glue on end grain and their projects are still holding together perfectly fine. But still, now that I know that glue works on end grain pretty well, I might as well add glue there as well. Unless it's a big glue up and I don't have enough time.
@@Mikey__R I find spreading glue on the end grain areas of the joint helps prevent cracking of the finish. A typical example would be the joinery of a cabinet door. I've seen many doors glued by just putting a bead on the tenon part and there is always a crack on the surface I. The finish. It is most noticeable when painted. With 100% coverage of glue throughout the joint, including on the end grain, this doesn't happen.
Excellent commentary. Practical application is always a trick. I’m not sure about using addition to calculate support. I can’t see a chair surviving the application of 2 tons. Plus there is uneven application of force. Thanks for the video!
Patrick's video is very important and relevant in the world of making beekeeping equipment. Everything we build is build for withdrawal strength of the fastener and pull apart strength of the joint. Nearly all of which have both end to side grain and side grain to side grain contact. I've been going down this avenue for awhile to find the best glue up method for building equipment as well as the most practical fastener (note not the strongest fastener...the most practical!). I'm amazed you got people emailing about getting rid of joinery in response to that video?! Seriously, who was the guy that thought, "oh, well I guess I can just butt joint everything now!"
Pretty easy to understand the issue... Woodworking joints take advantage of the strength of the long fibres, and have a significantly larger surface area. Glue joint strength is significantly dependent on surface area. Dovetail joints and box joints are very strong (someone said box joints were actually even stronger than dovetail joints, and this actually makes perfect sense when you understand the lignin issue). Joinery introduces frictional forces due to the tight fit, which increases the load needed to actually apply a shear load to the glue, making the joint unlikely to break. Then, the high load points in traditional joints resist motion along the fiber direction. If you understand compression and tension, you can understand a bend, where one side is in tension and the other is in compression. If the compression strength is higher than the tensile strength, it will fail at the end being stretched. The major traditional joints resist motion and torsion with the tension of wood fibers themselves, and the surface area of the glue is much higher, and in addition, other resisting loads are introduced. If you had a mortise and tenon, the way to break it with the least load would be to twist the tenon piece in a way that splits the mortise piece of wood, pulling the fibers apart laterally. Joint design has to reflect the expected loads. I.e. most loads are going to be vertical rather than lateral, and the lateral resistance to breaking might be significantly weaker on a joint. But this doesn't matter when the lateral loads are a fraction of the vertical loads. Think of how many pieces of furniture it's a bad idea to tip on its side and stand on, but how they can easily be stood on when loaded right-side-up. So many chairs will snap and break if you stand on them when they're laid out sideways, but are perfectly adequate for sitting.
Your comment near the end about the wood breaking partially due to the grain not being as strong in the long grain glue up direction brought it all together for me. Thx again and keep up the great work.
I gotta admit, I watched Stumpys video and was I influenced to watch & comment here. I was ready to defend Mr Sullivan’s findings after I saw the clip of you breaking the end grain joint of two longer pieces of wood over your knee.As You know Mr Sullivan did the same demonstration with the longer pieces but demonstrated both side grain and end grain joints with similar width pieces and proved the end grain to be more rigid. Realistically you are right , we woodworkers would never complete this joinery without tenoning or applying some type of doweling. Although Mr Sullivans video wasn’t a campaign for pro end grain joinery you expressed your opinion and supplemented his video without discrediting him and I appreciate that.
Very nice addition to Mr Sullivan's video. And I appreciate the civility of your response being a clarification of the original work; not a shouting down of results that don't fit the status quo. As you said, it's all information that leads to a better understanding of how we use wood and glue. That's science!
I'm glad you mentioned joints over time and how wood moves which weakens the joint. The flexibility of the glue to be able to compensate for this movement is also an important factor in the joints overall strength. Also that you pointed leverage multiplication!
I’m not going to change anything -but this information reassures me about the unreinforced miters that I use on my small boxes. I always thought (hoped) that splines were unnecessary. I now feel that there’s a new observation out there that says you don’t need splines for a joint that’s not going to be under a lot of stress.
Very valid and very timely point. This end glue strength thing just adds too, does not eliminate, joinery. It should be noted that Titebond has been recommending end-grain gluing for years and has had guidance on their tech site on the best way to do it. Go figure that the manufacturer was ignored by the industry they support???
I think you nailed the reason why the myth took hold in the first place. Glue breaks? The glue bond must be weak. Wood breaks? The glue bond must be strong. What we know now is the glue is stronger than lignin. And, thanks to your video, we are reminded that a tenon is MUCH stronger than a glue bond. Nice work!
I totally agree with what both Patrick Sullivan and you said. I might have highlighted why the long to end grain butt joint (the only one we actually really care about) will degrade over time with wood movement: the end grain part will expand and contract which will make the fibers connected by the glue on the two parts not match anymore which will make the joint weak (not an easy point to make in a short sentense).
This is a classic case of comparing apples to tomatoes. You made an excellent clarification and as an R&D manager who worked for major glue (Well adhesives that is), none of the end-users like furniture makes or structural wood products like GP, Wherehouser, etc. do the specific tests suited to the type of joints or glue-ups. Only very few scientific studies have been openly published where a comparison is done with end grain vs side grain joints under rigorous scientific conditions
I watched Patrick’s video the other day, thought it was very interesting. I agree with you that the real value in it was to be a bit more liberal with the glue on joints. I never was one to not bother end grain glue, this just confirmed it
The reason Patrick's video is important is that not everyone is making furniture or picture frames, with long pieces of wood and leverage trumping all. Guitar luthiers for example have small pieces of wood they use for reinforcement and have the option to choose end grain or long grain joints. But I like TWW's saying that yes, end grain joints are stronger, but we don't want to be in the cheap seats. We want to be closer to the wood fiber strength anyway, not glue joint strength.
Thanks for the follow-up. I'd heard about the video, but never watched it until this morning (because you had the actual link). I'd been suspecting something like Patrick's conclusion for a while, since I do a LOT of segmented turning. I was noticing that my "joints" were not failing as easily as conventional wisdom was claiming they should. But this does open up some options for me. But don't worry, I'm not getting rid of joinery for other projects. I hear what you're saying, Marc.
Thank you for this. I saw the Sullivan video, and thought to myself, there is more to it than this. I'm glad that you, and Stumpy Nums responded in a caring way to bring relevant context to it.
I enjoy seeing woodworkers share ideas and findings that they have made. Going along with how regular jointerie is better, you also briefly touched on wood movement. I think a lot if people are forgetting about the expansion and contraction that is going to break an end grain joint apart. That is the real reason that edge grain gluing is said to be stronger. It is not really about the strength for end and edge grain gluing it is about the longevity.
Fantastic explanation, dude! Thanks! 😃 But something that came to mind: mitered corners box vs a box made with rabbets. Shouldn't we reinforce both?! After all, the glue is going to hold better in the mitered box, because it's end grain to end grain. What do you think? Anyway, stay safe there with your family! 🖖😊
I know we are talking about PVA glue here.... But something about the strength of wood that amazed me was when a guy at my local Woodcraft store made a zigzag chair out of walnut using dominos and epoxy and they did a demo of the chair with three 200+ pound guys on that chair. And the walnut was 1”x2” if I remember correctly...
One more instance end grain-to-end grain strength is relevant, for us newbie woodworkers, is in more complex glueups where, say, rows of wood going into a cutting board are themselves made of several pieces of wood aligned end to end. It might be negligible to more advanced crafters, but till these videos came out I was actually quite worried that doing so would lead to weak connections promoting failure of the board. It's good to have such a simple matter clarified in an open way so a newcomer's progress and confidence isn't diminished by bad or old information.
Yup. Also, one might say something like "Well, whether it's 1200-1880 lbs or 4000 lbs, both of those are way stronger than the piece needs to be, so a glued butt joint is good enough for my needs". The loads that these videos discuss are essentially one-time static loads that bring the material/joint to its ultimate failure point. However, there is also the matter of fatigue strength. I know it works this way with metal - not certain about wood/glue joints - but fatigue occurs when a load is applied over many (hundreds, thousands, tens of thousands, or more) of cycles, and the cyclical load that results in failure may be only a fraction of the ultimate strength of a material. So maybe 1200-1880lbs of ultimate static strength may mean that years of loading at only a few hundred pounds might break the piece (maybe a chair like Marc said), while the 4000lb structure may have a fatigue strength of 1000lb or more. Wood movement across the joint, small as it may be, may also deteriorate the joint with time. I'll stick with the traditional joinery.
This was a good followup to Patrick's video. Patrick did debunk the myth, and Marc explained why the myth came about. It's all good knowledge. Edit: that article from Fine Woodworking is amazing. I highly recommend anyone who hasn't read it to do so. I found it at my local library as well as online.
1. I agree with pretty much everything you said 2. I agree that part of the idea of a weak joint comes from the clean break and leverage, i would also assume it comes from wood movement over time. Two pieces of wood, moving at different rates (unless they have the exact same properties) are going to weaken an end grain glue up vs. A side grain glue up that just pushes against each other.
Frankly, I think Patrick's video has the potential to affect how we think about three things. 1) Glue joints are not weaker just because end-grain is involved. 2) We probably don't need to spline our miters on picture frames. If the frame profile has a decent cross sectional area for the total weight, glue is strong enough. I would have said this 10 years ago, but nobody would have believed me. Some people still wont. 3) An edge-glued joint is stronger than the wood. Connecting edge joints with loose tenons doesn't final product stronger. If you want to use them for alignment, fine, but it doesn't generally make for a stronger panel.* Many of us are well aware of this, but since there are so many UA-cam videos of people using dominos in edge glue-ups, I thought it was worth mentioning. The force required to break a joint depends on many factors, so it's very hard to make apples-to-apples tests for different joinery techniques. It's really only fair to compare joinery techniques that share the same external geometry, because that's the only way the same forces will result in the same net stress on the joint. That's important because force doesn't determine when a joint fails - stress does. Stress comes in three flavors: tension (pulling) , compression (crushing), and shear (sliding). Different wood species and different glues will have different failure strengths for these three stresses. Furthermore, wood's ability withstand these stresses depends on whether the stress is applied longitudinally (along the grain), radially ("perpendicular" to the growth rings), or tangentially ("parallel" to the growth rings). Since wood rarely has grain that is perfectly oriented relative to the edges of our boards, and most joinery mixes grain directions, almost all real-world applications involve a mix of all these factors. Speaking generally, glue is 2x to 4x stronger in shear than in tension, and wood is 5x to 10x stronger in longitudinal tension than in radial or tangential shear. Then as a joint ages, it's affected by glue deterioration (admittedly slow with modern glues assuming you choose a suitable glue for the species and service conditions), wood shrinkage, and fatigue from intermittent loading. How a brand-new, well-fitted joint handles these stresses is likely different from how a well-used piece of furniture will handle them 15 years later. If you haven't figured it out yet, the structural analysis of wood joinery is complicated. Patrick's methodology effectively tests tensile stress in new, well-fitted joints for a variety of woods under both longitudinal and tangential loads using PVA glue. His video is informative and I feel smarter for having watched it, but it has very limited applications. What I hope I illustrated here is that few of us (myself included) understand wood joinery in anything approaching a quantifiable way. As such, we ignore centuries of hard-earned, traditional wisdom at our peril. *To be specific, a loose tenon won't increase the tensile or compressive strength (and, by extension, the bending strength) of the final panel. It probably increases the shear strength of the glue joint, but that's not how we normally load things like table tops and raised panels where edge glue-ups are most common, but there may be exceptions where shear is important.
Marc, I completely agree with your points about joinery but you also mentioned that we don’t glue boards together to make them longer, we get a longer board. I think the issue is, based on Patrick’s video, is could we? As Patrick explained, not for a narrow board, but we join boards on the edge grain all the time to make boards wider. Given sufficient width (i.e. glue surface), there appears to be no reason we could not reasonably join boards to makes them longer as well. If it failed, it would still fail along the joint, but the joint would be stronger than the failure point of our edge joined board. I’m not suggesting making end grain joined diving boards, but I can envision times where it might be useful that I may not have considered before. P.S., I’m still wondering about the wood movement variable….
I mean, you could…by why would you. You end up with a much weaker board and a glue line in the middle of a board. It’s definitely not ideal. Joining boards to make them wider is a different beast as it’s usually for table tops and panels, which won’t experiences them stresses of single long boards.
Mechanical Connections, Joinery, Pocket Hole Screw and Nail etc etc are enhanced by glue and glue is enhanced by Mechanical Connections. Excellent videos all round.
I think the other interesting take-away from Patrick's video is that with wide, short boards (e.g. book-matching resawn pieces into a longer panel), end-to-end glue joints are perhaps a valid option.
@@brendanloconnell sure, but if you did, you'd end up with board that's only twice as strong along the grain as it is across. A 3 to 4ft board with a weak point in the middle wouldn't be any use for anything structural, but you might find a purely decorative use for it? Maybe you could use those shorter boards to make frame and panel sides for a low bookcase?
Thank you for the calm and rational response to irrational ravings of the internet trolls and the grossly misinformed in the woodworking community. May cooler heads prevail! 😁👍👍
Excellent video and great clarification Marc. As you noted, real life joinery techniques are way different than what Mr. Sullivan was demonstrating in his testing procedures.
This was a good video, providing a lot of extra context to something Patrick stated in his his video about the strength of joints being related to the long grain running across the joint. Curiously though, last year while working on a shelving project using plywood, I had a situation whereby I might of done something slightly differently had I had the experience of Patrick’s video back then. I was edge gluing a couple of scraps to make a piece wide enough for a shelf. The joint itself did not need the strength of long grain crossing it, but because of my concern about the fact that each ply would be 90 degrees from its opposite number, i.e. edge grain to to end grain, and thus create a weak glue joint, I placed a bunch of dowels across the joint. Probably if I were doing the same thing in the future, I would just glue it and call it good.
I suppose that depends on the leverage forces that would be occurring in that scenario. If the shelves were long and poorly supported then a heavy object on either or both ends could cause failure. I'd have gone with the dowels, basically!
@@ricos1497 The shelf was not long and the joint was along its length as opposed to cutting across its with. In other words, it would be like a typical panel glue-up whereby you would use biscuits/dowels/dominos for alignment rather than strength.
Great insight and thanks for sharing your thoughts on it. I did not extrapolate his results to include joinery and saw no reason to. Loved you addition to the conversation.
Excellent supplement to Patrick's video. Great job. "You must add dowels, dominos, tenons, etc. to end grain joinery because glue alone is not strong enough." This generally true axiom somewhere became popularly misunderstood as "Glue adds ZERO strength to end grain so don't bother adding glue to end grain." Patrick's video yields the generally true axiom "End grain glue is stronger than long grain glue." Which will now be popularly misunderstood as "End grain glue is stronger than edge grain glue, so don't bother re-enforcing end grain joints."
"End grain glue is stronger than long grain glue" That's not what was tested or showcased though.... More like: "End grain glue is stronger than the long grain"
@@hakancarlsson2881 agreed I may have misstated that point and which has always been clear to many of us. My hope is that people who misunderstood existing end grain rules of thumb simply don’t swap old misunderstandings for new misunderstandings.
Yeah, I thought about that, seeing the original video. And there also is the matter of time. How will it fare over months or years. Having a good traditional joint, that sometimes in a dry fit is hard to break apart, only stacks the cards in your favour. And yeah, in order to get a like I stole that saying from you, figured it would stack the cards in my favour 😁
My gosh, I knew people would misunderstand Patrick's video or purposely twist it to cause controversy but some of the comments here and elsewhere exceed my expectations. Thank you for attempt to straighten out the confusion.
I left his video thinking-"hmm, that's good to know." Not, "I'm going to use end grain joints for everything! No dowels or biscuits or domino's needed. Only suckers use those!"
One video was a scientific explanation with virtually no practical use. This video is useful because it explains what woodworkers have known through experience for hundreds of years.
Have you tried the Beadlock? I bought one and tried it (thankfully on scraps) and it does not cut the slots parallel with the face of the wood causing the two pieces of wood at the joint to be out of plane. I tried several times to fix it but the joints will not be flush. And I made sure to reference the same face on both boards and to clamp in such a way that the jig didn’t shift when drilling (this happened easily on and end board when drilling in position B for the interior 2 holes). Now I’m stuck
The thing I would add is that joinery increases the surface area of the glue joint. The end grain to face grain might be slightly stronger than face to face, but if you've got four times as much surface area in the face to face joint guess which joint is stronger? That's where the strength of lap or box joints comes from, tons of surface area.
Two great video's. Sounds like the real myth is that glue is stronger than wood. Yes in some ways you measure those strength it is, BUT when it comes to woodworking a proper joint (of any type) is what gives you a practical product. And as a newbie and a techie thinking modern products should change the way we do things those old fashioned joins still have a very important place in the real world.
Don't get carried away. The glue is a large part of the strength of a joint. Even for dovetail joint there is the pull perpendicular to the tail. The structure is part of the strength; the glue is another, in most cases greater. Also important is the most likely direction of forces of stress that can cause failure.
Rob Cosmen also did the same test and was surprised by the results . No matter how you look at it, a little extra glue strength and peace of mind can't hurt .
As you pointed out, a lot of it has to do with leverage compared to joint size. The joint length of face grain and edge grain joints is considerable compared to the length of an end grain joint. The glue always has the same strength, as does the wood. The smaller the joint, the easier it is to compromise the joint, and the failure occurs at its weakest link. In face/edge grain, that link is in the lignin holding the grain together. In an end grain joint, that failure is the glue itself. When you further compare the leverage the jointing members have against the glue, the end grain wins again. While I will always continue to do joinery of some type, I will also continue to glue my face frame (edge to end) joints on top of doing joinery, because I have learned through the years that it adds quite a bit of strength, and also helps to hide the joint better. Mr. Sullivan's video isn't revolutionary to those of us that have been in the industry (or in the hobby) for any length of time, but it does highlight that gluing all your joints, no matter the orientation, with proper technique, is a best practice. Thank you Mark, for pointing out that good techniques, and proper joinery are still, and always will be the best practice in whispering to the wood...
Great vid! I was looking forward to your take on this subject. Thanks for interpreting this and providing some history. The only issue I have with your explanation is when you multiplied the strength of the joints x4. From an engineering standpoint, 4 joints would be statically indeterminate (only 3 equations, so can only solve for 3 unknowns without some serious mathematical gymnastics that are beyond my comprehension). But your point is still clear - MORE stronger joints is MUCH better! Thanks for all you do!
Late to the party - but I think a fair conclusion from Patrick Sullivan's video is that glue is 1/4 to 1/8 the strength of wood fibers, depending on the wood - which is still a fair amount stronger than lignin. When thinking of strength in woodworking, therefore, we should consider three things: grain direction, grain direction and grain direction.
To try to summarize Patrick’s video, for a given glued area (area of faces joined), end glued joints are as good as (or better than) side joins. This doesn’t mean they will be strong enough for your application.
A solid explanation, Marc. I think part of confusion comes from the term, "wood strength", which is wholly dependent upon which direction you're trying to break the wood. Along the grain, a board is less strong because the lignin bond is breaking apart, not the fibers themselves. ("Glue is stronger than the wood" is what we've always heard when talking about edge grain, but that's actually misleading. It's really only stronger than the lignin bond.) Trying to break a board across grain is much tougher because the fibers are much stronger. So the typical "glue strength test" of edge grain vs end grain is really comparing apples to oranges - you're actually testing wood fiber strength verses lignin bond strength. This may be totally wrong but it could be the GLUE joint strength on edge grain vs end grain is exactly the same. Not sure how you could test that specifically but it's an interesting topic. But no, I don't plan on selling my chisels or domino (if I had one), just yet. ;-)
Thanks for putting that vid into context. I felt a little skeptical of it, but I can be dogmatic at times, and you’re right-information is information. It’s just, I’d never trust a joint involving just end grain.
What I learned from all of this: glue resins interact with the lignum (the concern about glue interacting with fibers doesn’t matter); the geometry of the joint is still the critical factor if you want the yield strength of the joint to approach that of the material; and end-grain only joints are surprisingly strong. Torque matters, particularly when the performance delta is 1::2-3 v 1::8. I will look for ways to get more end-grain involvement, but I’m not going to seat the bottom of my tenons in a mortise just to get an additional gluing surface. I will glue the tops of my dovetails. Great videos all around, thanks!
I watched it last week and his outcomes were very surprising. I posted it to my local FB woodworking group. It surprised a lot of woodworkers. I think it was very well done.
was the surface area equalized in the glue only and mortise/spline joint? Much of the differences could be from more surface area involved in the mortise and spline joints.
Isn't the simple way of saying the whole thing something like: if you need a wider panel, you can glue side to side, and be safe in the knowledge that even if the panel is glued up, the lignin will fail (well) before the glue joint, so nobody will ever complain about the fact the panel is a glue up -- any failure will be in the wood, not in the joint. On the other hand, if you need a longer panel and glue endgrain to endgrain, that *will be failure point* -- because you have introduced an obvious failure point, should a failure actually occur everyone will say you are an idiot. Also, it is easier to find longer panels, which avoid the issue of endgrain to engrain glueups, than wider panels, so glueing side to side makes sense practically and structurally, whereas glueing endgrains does not.
Well said, I was trying to think of times when you can't realistically just find a board that is long enough for whatever you're building and the only situation I could think of was boat building.
@@WobblycogsUk that and timber framing. And still on those cases they use massive scarph joints, which again take advantage of the resistance of wood and not the glue.
I recently discovered that one of the 4 by 4 I bought in the woodshop was glued somewhere in the middle. However they cut two sawtooth patterns in both ends and glued them together.. Dunno if it as strong as the whole 4 by 4. Didn't notice any difference and prolly wouldn't have noticed if not for grain color.
Yep - and they would be right in saying it is your fault - cause it would be your fault. glueing endgrain to endgrain is a weak JOINT, just it is not the case that the glue in it is magically weaker but the joint it self is weak as it does not have the continuous fibres that make the wood strong along its grain. scarf joint or lap joint - those should have been compared to the simpel endgrain glueup - that is what the saying about endgrain-joints is all about.
Been doing middle age European style timber framing for years now, I have never thought to just glue two pieces together endgrain to endgrain because here all of this (side to side lignin will fail, end to end glue bond will fail and along the grain will take the most to break the wood- that is why joinery came to be) was a part of 6th grade woodworking class. To be honest in classical timber framing one really doesn´t need glue or metal fasteners as wooden pegs and the right types of joints with no glue in them will do all the hard labour of keeping the structure solid. Sorry if my terms are off, English is not my first language.
I didn't see the video but saw the people making the assumptions you did. Some said it tongue in cheek and I think a few believe it. As you point out here, it isn't data for wood working. Here is something you know, but maybe helpful to a novice. Wood is made of fibers. If you break it, it splinters everywhere. The fibers keep it together until they separate from each other and break at different places. To break a board, you have to separate the fibers. In martial arts, they break boards by exploiting the weakest angle to the fibers. When you cut a board, you weaken the fibers on the edges. When you glue it together, the glue bonds the already broken fibers, so when you break the joint, the fibers near the joint, not far away, break. The glue may not be as strong as the center of the board, but it is stronger than the edges. End to end tinsel strength, I can see that working. You have the ends of thousands of fibers gripping the glue and no leverage. The collective fibers are giving it strength. But, like you proved with your knee, the board didn't break. The glue broke because you separated the outer edges first and then all the rest gave way. I broke a bunch of but to edge glue joints today from older builds I abandoned or replaced. I used leverage and the glue almost always kept the fibers of the edge board. In over a dozen joints, only the edge of the board where it met the glue gave way. BTW, just to be clear, the joints were pocket hole joints that were glued. I removed the screws and reallocated the wood to my current project. I'm glad you made the video because it's good to see what the talk is all about. Thank you.
Great video to add on to Patrick’s findings. While I complete agree with what you had to say wouldn’t the thickness and width of the wood being used have an effect on the strength of the joint? I would be curious to see the results of the strength of the glue joint between two boards being glued in a cross form or a even in a half lap type joint. Again great video Marc.
To me the key point to take from both Patrick's video and this one is that lignin isn't all that strong. Don't rely on that if you are going to have a lot of demands put on a joint. The secondary point is that the glue is stronger than lignin. So properly done, an end-grain to end-grain joint will be pretty strong compared to what we were taught to expect. Now I want to see someone testing scarf joints. . . There's probably a video on that as well.
as a structural engineer, the next step to test this would be to do straight tensile tests... as the other video pointed out in the end to side tests, the stiffness contributed a lot. therefore, tensile cares very little about stiffness, at least compared to bending (planer sections remain plane assumption). Though we'd probably still find the same thing - it was a statically determinant beam after all. just want to see a test trying to eliminate stiffness as a factor and truly test the glue bonding to the different types.
Love the shirt! I really like her videos and creations. But back to the subject at hand, I think you did a very good response and reaction video, showing how even though he isn’t wrong, we still strive for stronger joints than what only glue can provide.
Secretly I've been hoping you would make a response video on this. So thank you. Excellent explanation.
Insert the young Robert Redford nodding meme gif
Same!
Just because he said the words wood movement, doesn't mean you were right. We will continue this debate on Tuesday.
Crap, I had a video about this planned for today... Oh well, at least folks will stop emailing me asking for a response! Thanks, Marc!
We still want your take!
The early bird gets the worm but you should still do a video on this subject
@@DanDDirges hell yes! I didn’t even know that I wanted Marc’s take after seeing Patrick’s video, but here I am... the challenge, however, is adding something to the topic rather than just doing another ‘version’. Good luck!
IMHO, James, your voice would still be a valuable addition to the conversation.
Go for it Stumpy, your the authority...
If you haven't already, please go to Patrick’s video and give it a watch BEFORE watching this. ua-cam.com/video/m7HxBa9WVis/v-deo.html
This is the Internet. How dare you expect us to go watch the original video before making up our minds! 😏
Common sense applied, love it!
I wanted to see if you could break the end grain joint with a domino in it.
9:32 Marc: "I'm not the strongest human being out there".
Stumpy Nubs: "I've never seen him and Superman in the same room at the same time".
Patrick did a great job on his video. His channel is a real hidden gem among woodworking channels.
True words
Very cool to see Patrick shared on a bigger channel. Hope his audience grows!
Me too. I like his style.
Going on 50 yrs ago when I was a teen learning basic woodworking, I discovered that it actually DID help to glue things like the end grain shoulder of a tenon, or the end grain of miters, etc. More recently it’s pocket hole face frame joinery and the like.NOT because it necessarily makes anything “stronger” overall, but because it DOES keep all the joinery looking better throughout expansion/contraction cycles and typical daily use, to the point of having far fewer cosmetic issues like hairline cracks in finishes or dirt in the air discoloring the unglued edges over time. Your glue is already out so It takes a couple extra seconds… why NOT do it?
No no, please throw away your dominoes. Just tell me where, I'll come make sure they get recycled.
I watched Patrick's video a couple of days ago and until you mentioned it I did not think that anyone would think that you could do away with mortise and tendons. I was wrong again. Great points as usual Marc.
Sullivans Video ist very good. So is your explanation. Thanks. As an furniture restorer i see every day what joint fails. And it is the endgrain joint and the wiggly/loose joint. Never rely on glue. The Elder secured every joint with some sort of mechanical interlock like pegs or dovetails for a reason.
Great video and a shining example of respectfully adding to someone’s content. I really appreciate that you took the time to clarify at the beginning and end of the video that you agreed with him. You are saying that the myth is a fallacy, but it is still of little practical use. I also find it somewhat annoying that you have to make it that clear so people don’t twist your words.
No need to twist his words--he downplays Patrick Sullivan's remarkable result. He said Patrick was right BUT who cares. Instead he should have said Wow Patrick has flipped a standard claim of 100 years AND here is some clarification for those few people who may have misunderstood it's impact to joinery.
Man I'm glad you came out with this video before trash day! Was able to get my DF500 and 700XL out before pickup!
Seriously though I enjoyed his well made video (and his others) and was interesting to see the breakages along grain lines but it changed nothing about how I would approach any joinery.
Loved this video. I was thinking about wood movement the whole time end grain to edge grain was discussed. It would be interesting to see those tests after the wood has gone through a year of humidity changes.
could simulate fairly easily going through some humidifier/dehumidifier cycles
Excellent couple of videos by both TWW and Patrick. I came to UA-cam trying to figure out a joint I’m trying to make so the timing of this is excellent. If I use my Domino to make a joint where one end of the Domino sits in the mortise such that the long grain of the domino mates to the end grain on mating piece. The joint I’m trying to describe is like joining the sides to the back of a box. Would that work? Patricks video seems to imply that it would? I really don’t want to use screws or pocket holes, but there is no other way to orient the domino for this particular joint.
I was surprised by Mr Sullivan's conclusion when I watched his video. Then, you bring your knowledge and, applied to real world woodworking, everything become obvious: mechanical strength! Thanks Marc, it's your turn to make me feel more intelligent 😂!
Great follow up video. I agree everything Patrick said was correct but it doesn't really reflect what we generally do in woodworking - although perhaps it indicates we should change some of the things we do. The one thing that I'd like to have seen mentioned here though is the way the joint fails. With an end grain to end grain join the failure is sudden and catastrophic. Long grain failures, in my experience, usually give you a bit of warning before they let go and mechanical joints tend to give quite a lot of warning.
Thank you! This was my precise reaction to the video and some of its responses - albeit better said and available to a much wider audience. His findings completely (and somewhat shockingly) bust the myth. What they don't do is invalidate the hard-won lessons of hundreds of years of furniture makers vis-a-vis functional joinery. One reason the myth has stood for so long is that there is almost never any good reason to glue end grain to end grain - if you want to stretch a board you'll typically use some form of scarf joint which increases the glue surface and hence the strength.
The myth does make sense in the context of, if you're gluing side-grain to side-grain, you end up with a joint that is stronger than the rest of the board, and if you're gluing end-grain to end-grain, you end up with a join that is 1/4 to 1/8 as strong as the rest of the board. I don't think most people realize how relatively weak wood is if you stress it with the grain.
This is a point that only us structural Engineers probably appreciate---- Stress or pressure is not applied to a joint----- A. Force is applied to a joint----- Stress and pressure are the result of that force., and are highly dependent on the area that the force is spread over.
In joining end grain to side grain, apply a thin coat of Titebond to the end grain surface and let it dry before applying the glue to the two boards to be joined. This seals the end grain so it won't suck the glue away and starve the joint.
Completely agree. When using titebond 3 I change it up a little by not allowing the primer/sealer coat to fully dry. The additional dab of glue before assembly can then make a better chemical bond.
Patrick was just busting the end grain myth. The real takeaway is that if you would be OK with a joint using edge grain, there's no reason to think it wouldn't be OK with end grain. All the other things you would normally consider -- grain direction, wood movement, need for hardware, etc. -- still apply.
To add a minor thing, though -- the direction of the grain is also important when considering how hardware such as screws are used. Screws into end grain are not as weak as people often assume, but it can still be a bad choice depending on the application.
Screws into end grain tests - viral video idea? 😁
"The real takeaway is that if you would be OK with a joint using edge grain, there's no reason to think it wouldn't be OK with end grain."
This nicely encapsulated the important lessons. If you took a square frame cut out of solid wood and subjected it to load, you'd expect the failure to occur along the short sections of the grain. So yes, glue is stronger than wood, but wood itself isn't strong enough for the job if you're relying on what amounts to "end grain boards".
Also absent in must of this discussion is leverage. Sure, 400 lbs sounds like a lot, but that load is trivially reached by the mass of an adult placed out on the end of a lever. The joints of a chair in which someone is leaning backwards is subjected to many multiples of their weight.
I think it was an excellent video by Patrick, however my biggest gripe is that it's not really comparing the strength of a glue joint along the grain versus a glue joint with end grain. It's only really proving that the glue along the grain is stronger than the wood fibers themselves, as the wood is what broke in every test, as opposed in the end grain tests it was the actual glue joint that broke. I think that confuses a lot of people as well, even though that one of the tests do not show a failure of the glue joint, quite the opposite.
He did say he was willing to do additional testing if there was enough interest in it...
No, it did not show that the glue is stronger than the wood fibres! Quite the opposite, in fact. What his tests showed was that the glue is stronger than the LIGNIN holding the fibres together, but not as strong as the fibres. This is why long grain glueing broke the wood (ie the lignin bonds), and end grain glueing broke the glue. It took more force to break the end grain glueing exactly because the glue was stronger than the lignin.
Since the side grain glue joints all failed AWAY from the joint the only conclusion you can draw is that this glue Joint is at least as strong as the cross grain strength of the wood. Because it never reached failure in the glue joint, you cannot say that it is weaker than an end grain glue joint
@@jeffhigh2 No, this is not entirely true. Let me try and explain what/how Patrick Sullivan (PS) measures:
Initially, there are two variables: glue strength and wood strength. However, for this purpose, we can break down wood into two sub-components, fibres and lignin:
Fibres are cellulose tubes that are the transportation means for water and nutrients in the tree. Lignin is the glue that holds these fibres together.
Thus, we can talk about glue strength, fibre strength, and lignin strength:
Glue strength is measured as the force it takes to break two pieces of wood along a glue joint (S^g). 'Glue' is here a pva-glue. Fibre strength is measured as the force it takes to break a piece of wood across the fibres (S^f). Lignin strength is measured as the force it takes to break the fibres apart from each other, ie break a piece of wood along the fibres (S^l).
From the empirical tests performed by PS (watch the video!), the only general information we can infer, is that glue is stronger than lignin (along grain glue-ups are bound to break away from the glue line):
(S^g) > (S^l)
End grain to end grain glue ups do not provide any relative information about strength even though they are bound to break at the glue line, since the wood fibres have already been severed:
* => (S^f) > (S^g)
Thus, end grain glue ups provide glue strength information only.
Luckily, PS does not rely on relative information only, but he also performs absolute tests: he measures how much force it takes to break apart two given test pieces. He tests a number of wood species for strength in different directions (Sw^f and Sw^l) as well as glue strength for the given wood species, tested as end grain glue ups (Sw^g).
These tests all reveal that:
Sw^f > Sw^g > Sw^l
For all tested species and specimens, fibres are stronger than glue, and glue is stronger than lignin. The actual force required differs from wood species to wood species, however all tests show the same relative order, namely that end grain glue ups require more force to break than along grain glue ups. This is what is illustrated in the graphs also shown in this video.
Remember, these are test tube results, and they do not tell us anything about if or how we should change the way we work, as is the gist of this video, but they DO provide us with a piece of information that is most likely new to most of us. Bravo Patrick!
@@pinkerbot But the test never actually breaks a side grain joint, the material breaks before the glue fails. You have effectively just measured the cross grain bending strength of the material (Which is enhanced at the joint by glue penetration) What the test tells us is that an end grain glue joint can be as strong as the wood cross grain in bending (glued or not)
Marc you now MUST connect those 2 pieces with a M&T and try and break them.
hello, What i'm missing here is that when you bring a piece of furniture to a client/ buyer you want it to be as strong as possible because when it breaks or get damaged because it's made to weak you have to go back an fix it, you never have time for that and the client is sort of disappointed. So my point is: engineer the piece in a certain way that it's durable. It's already pricy when you make handbuild stuff so expectations are higher. Still it is really good to know all this background information about the strength etc.. really helpfull and gives you a bit of confidence when you make stuff and you think yeah it's durable enough.
greetings from Europe
Great job adding context to this topic. I loved you highlighting your concern with what viewers would take away from Patrick’s video, not his video per se. The good YTers realize it’s virtually impossible to have the message a creator intends to convey be the one received by all that watch it.
He does mention that a future test will be on mechanical joints. I’m looking forward to that one.
Thank god someone has some common sense. People are acting like this video destroyed centuries of wood working techniques and wood workers are doing it wrong.
Great video Marc. We've been inundated with questions on this lately and like you, I can't really come up with a real-world use for an end grain joint-save one, on the podcast we answered a question about solid-wood chess boards and Patrick's video would make me rethink my answer on that one. -Ben
Oh... that article is in front of the paywall!
Another end grain to end grain use... well... kind of the same one, create a pattern. ua-cam.com/video/NUTmBy-63qo/v-deo.html
My only real conclusion from Patrick's video was, if my glue up allows the time for it, I will now add glue to end grain in addition to just long grain to long grain contact surfaces.
@@edu_carceller I'm not sure that's necessary. We've all collectively been building furniture using white and yellow glue for decades now, and for the most part if the joinery is sound, the furniture holds together just fine. I'm not convinced adding a little extra glue strength is going to make much difference, when the bulk of the strength comes from the joinery itself.
@@Mikey__R Oh I am sure it is not necessary because tons of woodworkers of all skill levels have been avoiding glue on end grain and their projects are still holding together perfectly fine. But still, now that I know that glue works on end grain pretty well, I might as well add glue there as well. Unless it's a big glue up and I don't have enough time.
@@Mikey__R I find spreading glue on the end grain areas of the joint helps prevent cracking of the finish. A typical example would be the joinery of a cabinet door. I've seen many doors glued by just putting a bead on the tenon part and there is always a crack on the surface I. The finish. It is most noticeable when painted. With 100% coverage of glue throughout the joint, including on the end grain, this doesn't happen.
Excellent commentary. Practical application is always a trick. I’m not sure about using addition to calculate support. I can’t see a chair surviving the application of 2 tons. Plus there is uneven application of force. Thanks for the video!
Patrick's video is very important and relevant in the world of making beekeeping equipment. Everything we build is build for withdrawal strength of the fastener and pull apart strength of the joint. Nearly all of which have both end to side grain and side grain to side grain contact. I've been going down this avenue for awhile to find the best glue up method for building equipment as well as the most practical fastener (note not the strongest fastener...the most practical!). I'm amazed you got people emailing about getting rid of joinery in response to that video?! Seriously, who was the guy that thought, "oh, well I guess I can just butt joint everything now!"
Pretty easy to understand the issue...
Woodworking joints take advantage of the strength of the long fibres, and have a significantly larger surface area.
Glue joint strength is significantly dependent on surface area.
Dovetail joints and box joints are very strong (someone said box joints were actually even stronger than dovetail joints, and this actually makes perfect sense when you understand the lignin issue).
Joinery introduces frictional forces due to the tight fit, which increases the load needed to actually apply a shear load to the glue, making the joint unlikely to break. Then, the high load points in traditional joints resist motion along the fiber direction.
If you understand compression and tension, you can understand a bend, where one side is in tension and the other is in compression. If the compression strength is higher than the tensile strength, it will fail at the end being stretched.
The major traditional joints resist motion and torsion with the tension of wood fibers themselves, and the surface area of the glue is much higher, and in addition, other resisting loads are introduced.
If you had a mortise and tenon, the way to break it with the least load would be to twist the tenon piece in a way that splits the mortise piece of wood, pulling the fibers apart laterally. Joint design has to reflect the expected loads. I.e. most loads are going to be vertical rather than lateral, and the lateral resistance to breaking might be significantly weaker on a joint.
But this doesn't matter when the lateral loads are a fraction of the vertical loads.
Think of how many pieces of furniture it's a bad idea to tip on its side and stand on, but how they can easily be stood on when loaded right-side-up. So many chairs will snap and break if you stand on them when they're laid out sideways, but are perfectly adequate for sitting.
Your comment near the end about the wood breaking partially due to the grain not being as strong in the long grain glue up direction brought it all together for me. Thx again and keep up the great work.
I gotta admit, I watched Stumpys video and was I influenced to watch & comment here. I was ready to defend Mr Sullivan’s findings after I saw the clip of you breaking the end grain joint of two longer pieces of wood over your knee.As You know Mr Sullivan did the same demonstration with the longer pieces but demonstrated both side grain and end grain joints with similar width pieces and proved the end grain to be more rigid. Realistically you are right , we woodworkers would never complete this joinery without tenoning or applying some type of doweling.
Although Mr Sullivans video wasn’t a campaign for pro end grain joinery you
expressed your opinion and supplemented his video without discrediting him and I appreciate that.
The intro was perfect! 😘
Very nice addition to Mr Sullivan's video. And I appreciate the civility of your response being a clarification of the original work; not a shouting down of results that don't fit the status quo. As you said, it's all information that leads to a better understanding of how we use wood and glue. That's science!
I'm glad you mentioned joints over time and how wood moves which weakens the joint. The flexibility of the glue to be able to compensate for this movement is also an important factor in the joints overall strength. Also that you pointed leverage multiplication!
I’m not going to change anything -but this information reassures me about the unreinforced miters that I use on my small boxes. I always thought (hoped) that splines were unnecessary. I now feel that there’s a new observation out there that says you don’t need splines for a joint that’s not going to be under a lot of stress.
A good joint transfers the load into the rest of the structure.
I was screaming at my phone watching Patricks videos. Thank you for making an civillized version of that.
Very valid and very timely point. This end glue strength thing just adds too, does not eliminate, joinery. It should be noted that Titebond has been recommending end-grain gluing for years and has had guidance on their tech site on the best way to do it. Go figure that the manufacturer was ignored by the industry they support???
I think you nailed the reason why the myth took hold in the first place. Glue breaks? The glue bond must be weak. Wood breaks? The glue bond must be strong. What we know now is the glue is stronger than lignin. And, thanks to your video, we are reminded that a tenon is MUCH stronger than a glue bond. Nice work!
I totally agree with what both
Patrick Sullivan and you said.
I might have highlighted why the long to end grain butt joint (the only one we actually really care about) will degrade over time with wood movement: the end grain part will expand and contract which will make the fibers connected by the glue on the two parts not match anymore which will make the joint weak (not an easy point to make in a short sentense).
This is a classic case of comparing apples to tomatoes. You made an excellent clarification and as an R&D manager who worked for major glue (Well adhesives that is), none of the end-users like furniture makes or structural wood products like GP, Wherehouser, etc. do the specific tests suited to the type of joints or glue-ups. Only very few scientific studies have been openly published where a comparison is done with end grain vs side grain joints under rigorous scientific conditions
Well.. the MAIN thing I learned here, is that we're able to buy *Ashley Harwood* shirts! Now I want one.
Me too, looks so cool
Great explanation for people to understand how they should understand what was being presented.
I watched Patrick’s video the other day, thought it was very interesting. I agree with you that the real value in it was to be a bit more liberal with the glue on joints. I never was one to not bother end grain glue, this just confirmed it
As usual great information. Been watching for 7 years now , thanks.
The reason Patrick's video is important is that not everyone is making furniture or picture frames, with long pieces of wood and leverage trumping all. Guitar luthiers for example have small pieces of wood they use for reinforcement and have the option to choose end grain or long grain joints.
But I like TWW's saying that yes, end grain joints are stronger, but we don't want to be in the cheap seats. We want to be closer to the wood fiber strength anyway, not glue joint strength.
Thanks for the follow-up. I'd heard about the video, but never watched it until this morning (because you had the actual link). I'd been suspecting something like Patrick's conclusion for a while, since I do a LOT of segmented turning. I was noticing that my "joints" were not failing as easily as conventional wisdom was claiming they should. But this does open up some options for me.
But don't worry, I'm not getting rid of joinery for other projects. I hear what you're saying, Marc.
Thank you for this. I saw the Sullivan video, and thought to myself, there is more to it than this. I'm glad that you, and Stumpy Nums responded in a caring way to bring relevant context to it.
I enjoy seeing woodworkers share ideas and findings that they have made. Going along with how regular jointerie is better, you also briefly touched on wood movement. I think a lot if people are forgetting about the expansion and contraction that is going to break an end grain joint apart. That is the real reason that edge grain gluing is said to be stronger. It is not really about the strength for end and edge grain gluing it is about the longevity.
Fantastic explanation, dude! Thanks! 😃
But something that came to mind: mitered corners box vs a box made with rabbets. Shouldn't we reinforce both?! After all, the glue is going to hold better in the mitered box, because it's end grain to end grain.
What do you think?
Anyway, stay safe there with your family! 🖖😊
I know we are talking about PVA glue here.... But something about the strength of wood that amazed me was when a guy at my local Woodcraft store made a zigzag chair out of walnut using dominos and epoxy and they did a demo of the chair with three 200+ pound guys on that chair. And the walnut was 1”x2” if I remember correctly...
One more instance end grain-to-end grain strength is relevant, for us newbie woodworkers, is in more complex glueups where, say, rows of wood going into a cutting board are themselves made of several pieces of wood aligned end to end. It might be negligible to more advanced crafters, but till these videos came out I was actually quite worried that doing so would lead to weak connections promoting failure of the board. It's good to have such a simple matter clarified in an open way so a newcomer's progress and confidence isn't diminished by bad or old information.
Wery well done!!! I tried to think of many ways to explain it and everyone was confusing.
Yup. Also, one might say something like "Well, whether it's 1200-1880 lbs or 4000 lbs, both of those are way stronger than the piece needs to be, so a glued butt joint is good enough for my needs". The loads that these videos discuss are essentially one-time static loads that bring the material/joint to its ultimate failure point. However, there is also the matter of fatigue strength. I know it works this way with metal - not certain about wood/glue joints - but fatigue occurs when a load is applied over many (hundreds, thousands, tens of thousands, or more) of cycles, and the cyclical load that results in failure may be only a fraction of the ultimate strength of a material. So maybe 1200-1880lbs of ultimate static strength may mean that years of loading at only a few hundred pounds might break the piece (maybe a chair like Marc said), while the 4000lb structure may have a fatigue strength of 1000lb or more. Wood movement across the joint, small as it may be, may also deteriorate the joint with time. I'll stick with the traditional joinery.
...and knowing is half the battle. G.I. Marc!!!
I love this community! Patrick's video was great, and you all bring a level of camaraderie to this craft by building on each other's work!
This was a good followup to Patrick's video. Patrick did debunk the myth, and Marc explained why the myth came about. It's all good knowledge.
Edit: that article from Fine Woodworking is amazing. I highly recommend anyone who hasn't read it to do so. I found it at my local library as well as online.
1. I agree with pretty much everything you said
2. I agree that part of the idea of a weak joint comes from the clean break and leverage, i would also assume it comes from wood movement over time. Two pieces of wood, moving at different rates (unless they have the exact same properties) are going to weaken an end grain glue up vs. A side grain glue up that just pushes against each other.
Frankly, I think Patrick's video has the potential to affect how we think about three things. 1) Glue joints are not weaker just because end-grain is involved. 2) We probably don't need to spline our miters on picture frames. If the frame profile has a decent cross sectional area for the total weight, glue is strong enough. I would have said this 10 years ago, but nobody would have believed me. Some people still wont. 3) An edge-glued joint is stronger than the wood. Connecting edge joints with loose tenons doesn't final product stronger. If you want to use them for alignment, fine, but it doesn't generally make for a stronger panel.* Many of us are well aware of this, but since there are so many UA-cam videos of people using dominos in edge glue-ups, I thought it was worth mentioning.
The force required to break a joint depends on many factors, so it's very hard to make apples-to-apples tests for different joinery techniques. It's really only fair to compare joinery techniques that share the same external geometry, because that's the only way the same forces will result in the same net stress on the joint. That's important because force doesn't determine when a joint fails - stress does.
Stress comes in three flavors: tension (pulling) , compression (crushing), and shear (sliding). Different wood species and different glues will have different failure strengths for these three stresses. Furthermore, wood's ability withstand these stresses depends on whether the stress is applied longitudinally (along the grain), radially ("perpendicular" to the growth rings), or tangentially ("parallel" to the growth rings). Since wood rarely has grain that is perfectly oriented relative to the edges of our boards, and most joinery mixes grain directions, almost all real-world applications involve a mix of all these factors. Speaking generally, glue is 2x to 4x stronger in shear than in tension, and wood is 5x to 10x stronger in longitudinal tension than in radial or tangential shear. Then as a joint ages, it's affected by glue deterioration (admittedly slow with modern glues assuming you choose a suitable glue for the species and service conditions), wood shrinkage, and fatigue from intermittent loading. How a brand-new, well-fitted joint handles these stresses is likely different from how a well-used piece of furniture will handle them 15 years later.
If you haven't figured it out yet, the structural analysis of wood joinery is complicated. Patrick's methodology effectively tests tensile stress in new, well-fitted joints for a variety of woods under both longitudinal and tangential loads using PVA glue. His video is informative and I feel smarter for having watched it, but it has very limited applications. What I hope I illustrated here is that few of us (myself included) understand wood joinery in anything approaching a quantifiable way. As such, we ignore centuries of hard-earned, traditional wisdom at our peril.
*To be specific, a loose tenon won't increase the tensile or compressive strength (and, by extension, the bending strength) of the final panel. It probably increases the shear strength of the glue joint, but that's not how we normally load things like table tops and raised panels where edge glue-ups are most common, but there may be exceptions where shear is important.
Marc, I completely agree with your points about joinery but you also mentioned that we don’t glue boards together to make them longer, we get a longer board. I think the issue is, based on Patrick’s video, is could we? As Patrick explained, not for a narrow board, but we join boards on the edge grain all the time to make boards wider. Given sufficient width (i.e. glue surface), there appears to be no reason we could not reasonably join boards to makes them longer as well. If it failed, it would still fail along the joint, but the joint would be stronger than the failure point of our edge joined board. I’m not suggesting making end grain joined diving boards, but I can envision times where it might be useful that I may not have considered before. P.S., I’m still wondering about the wood movement variable….
I mean, you could…by why would you. You end up with a much weaker board and a glue line in the middle of a board. It’s definitely not ideal. Joining boards to make them wider is a different beast as it’s usually for table tops and panels, which won’t experiences them stresses of single long boards.
Mechanical Connections, Joinery, Pocket Hole Screw and Nail etc etc are enhanced by glue and glue is enhanced by Mechanical Connections. Excellent videos all round.
I think the other interesting take-away from Patrick's video is that with wide, short boards (e.g. book-matching resawn pieces into a longer panel), end-to-end glue joints are perhaps a valid option.
Maybe, but I think if it were me, I'd use a continuous board as a substrate, and use bookmatched veneers to make it pretty.
@@Mikey__R I'm thinking more in the case of milling firewood length pieces (
@@brendanloconnell sure, but if you did, you'd end up with board that's only twice as strong along the grain as it is across. A 3 to 4ft board with a weak point in the middle wouldn't be any use for anything structural, but you might find a purely decorative use for it?
Maybe you could use those shorter boards to make frame and panel sides for a low bookcase?
Thank you for the calm and rational response to irrational ravings of the internet trolls and the grossly misinformed in the woodworking community. May cooler heads prevail! 😁👍👍
Excellent video and great clarification Marc. As you noted, real life joinery techniques are way different than what Mr. Sullivan was demonstrating in his testing procedures.
This was a good video, providing a lot of extra context to something Patrick stated in his his video about the strength of joints being related to the long grain running across the joint.
Curiously though, last year while working on a shelving project using plywood, I had a situation whereby I might of done something slightly differently had I had the experience of Patrick’s video back then. I was edge gluing a couple of scraps to make a piece wide enough for a shelf. The joint itself did not need the strength of long grain crossing it, but because of my concern about the fact that each ply would be 90 degrees from its opposite number, i.e. edge grain to to end grain, and thus create a weak glue joint, I placed a bunch of dowels across the joint. Probably if I were doing the same thing in the future, I would just glue it and call it good.
I suppose that depends on the leverage forces that would be occurring in that scenario. If the shelves were long and poorly supported then a heavy object on either or both ends could cause failure. I'd have gone with the dowels, basically!
@@ricos1497 The shelf was not long and the joint was along its length as opposed to cutting across its with. In other words, it would be like a typical panel glue-up whereby you would use biscuits/dowels/dominos for alignment rather than strength.
Excellent video, I love your conclusion. Information is always good, the best we know the best we can archive
Great insight and thanks for sharing your thoughts on it. I did not extrapolate his results to include joinery and saw no reason to. Loved you addition to the conversation.
Excellent supplement to Patrick's video. Great job.
"You must add dowels, dominos, tenons, etc. to end grain joinery because glue alone is not strong enough." This generally true axiom somewhere became popularly misunderstood as "Glue adds ZERO strength to end grain so don't bother adding glue to end grain."
Patrick's video yields the generally true axiom "End grain glue is stronger than long grain glue." Which will now be popularly misunderstood as "End grain glue is stronger than edge grain glue, so don't bother re-enforcing end grain joints."
"End grain glue is stronger than long grain glue"
That's not what was tested or showcased though....
More like:
"End grain glue is stronger than the long grain"
@@hakancarlsson2881 agreed I may have misstated that point and which has always been clear to many of us. My hope is that people who misunderstood existing end grain rules of thumb simply don’t swap old misunderstandings for new misunderstandings.
@@woodshopnerdery 👍
Yeah, I thought about that, seeing the original video. And there also is the matter of time. How will it fare over months or years.
Having a good traditional joint, that sometimes in a dry fit is hard to break apart, only stacks the cards in your favour.
And yeah, in order to get a like I stole that saying from you, figured it would stack the cards in my favour 😁
I think we’ve all seen that video. Good info all around and thanks for the additional points
My gosh, I knew people would misunderstand Patrick's video or purposely twist it to cause controversy but some of the comments here and elsewhere exceed my expectations. Thank you for attempt to straighten out the confusion.
I left his video thinking-"hmm, that's good to know." Not, "I'm going to use end grain joints for everything! No dowels or biscuits or domino's needed. Only suckers use those!"
One video was a scientific explanation with virtually no practical use. This video is useful because it explains what woodworkers have known through experience for hundreds of years.
You are absolutely right, no glue alone joint will ever beat a properly executed mortise and tenon joint.
Have you tried the Beadlock? I bought one and tried it (thankfully on scraps) and it does not cut the slots parallel with the face of the wood causing the two pieces of wood at the joint to be out of plane. I tried several times to fix it but the joints will not be flush. And I made sure to reference the same face on both boards and to clamp in such a way that the jig didn’t shift when drilling (this happened easily on and end board when drilling in position B for the interior 2 holes). Now I’m stuck
The thing I would add is that joinery increases the surface area of the glue joint. The end grain to face grain might be slightly stronger than face to face, but if you've got four times as much surface area in the face to face joint guess which joint is stronger? That's where the strength of lap or box joints comes from, tons of surface area.
I think you covered my concerns(moisture, time, mechanical) with the original video. So, interesting, but no change required.
Thank you! This video came as I’m about to do a project where I need to do an end grain glue up.
Two great video's. Sounds like the real myth is that glue is stronger than wood. Yes in some ways you measure those strength it is, BUT when it comes to woodworking a proper joint (of any type) is what gives you a practical product.
And as a newbie and a techie thinking modern products should change the way we do things those old fashioned joins still have a very important place in the real world.
Don't get carried away. The glue is a large part of the strength of a joint. Even for dovetail joint there is the pull perpendicular to the tail.
The structure is part of the strength; the glue is another, in most cases greater. Also important is the most likely direction of forces of stress that can cause failure.
Rob Cosmen also did the same test and was surprised by the results . No matter how you look at it, a little extra glue strength and peace of mind can't hurt .
As you pointed out, a lot of it has to do with leverage compared to joint size. The joint length of face grain and edge grain joints is considerable compared to the length of an end grain joint. The glue always has the same strength, as does the wood. The smaller the joint, the easier it is to compromise the joint, and the failure occurs at its weakest link. In face/edge grain, that link is in the lignin holding the grain together. In an end grain joint, that failure is the glue itself.
When you further compare the leverage the jointing members have against the glue, the end grain wins again.
While I will always continue to do joinery of some type, I will also continue to glue my face frame (edge to end) joints on top of doing joinery, because I have learned through the years that it adds quite a bit of strength, and also helps to hide the joint better.
Mr. Sullivan's video isn't revolutionary to those of us that have been in the industry (or in the hobby) for any length of time, but it does highlight that gluing all your joints, no matter the orientation, with proper technique, is a best practice.
Thank you Mark, for pointing out that good techniques, and proper joinery are still, and always will be the best practice in whispering to the wood...
Great vid! I was looking forward to your take on this subject. Thanks for interpreting this and providing some history.
The only issue I have with your explanation is when you multiplied the strength of the joints x4. From an engineering standpoint, 4 joints would be statically indeterminate (only 3 equations, so can only solve for 3 unknowns without some serious mathematical gymnastics that are beyond my comprehension). But your point is still clear - MORE stronger joints is MUCH better!
Thanks for all you do!
Late to the party - but I think a fair conclusion from Patrick Sullivan's video is that glue is 1/4 to 1/8 the strength of wood fibers, depending on the wood - which is still a fair amount stronger than lignin. When thinking of strength in woodworking, therefore, we should consider three things: grain direction, grain direction and grain direction.
To try to summarize Patrick’s video, for a given glued area (area of faces joined), end glued joints are as good as (or better than) side joins. This doesn’t mean they will be strong enough for your application.
A solid explanation, Marc. I think part of confusion comes from the term, "wood strength", which is wholly dependent upon which direction you're trying to break the wood. Along the grain, a board is less strong because the lignin bond is breaking apart, not the fibers themselves. ("Glue is stronger than the wood" is what we've always heard when talking about edge grain, but that's actually misleading. It's really only stronger than the lignin bond.) Trying to break a board across grain is much tougher because the fibers are much stronger. So the typical "glue strength test" of edge grain vs end grain is really comparing apples to oranges - you're actually testing wood fiber strength verses lignin bond strength. This may be totally wrong but it could be the GLUE joint strength on edge grain vs end grain is exactly the same. Not sure how you could test that specifically but it's an interesting topic.
But no, I don't plan on selling my chisels or domino (if I had one), just yet. ;-)
Thank you.
Both for pointing me at Patrick's excellent video and for adding your own "spin" to it.
Thanks for putting that vid into context. I felt a little skeptical of it, but I can be dogmatic at times, and you’re right-information is information. It’s just, I’d never trust a joint involving just end grain.
What I learned from all of this: glue resins interact with the lignum (the concern about glue interacting with fibers doesn’t matter); the geometry of the joint is still the critical factor if you want the yield strength of the joint to approach that of the material; and end-grain only joints are surprisingly strong. Torque matters, particularly when the performance delta is 1::2-3 v 1::8. I will look for ways to get more end-grain involvement, but I’m not going to seat the bottom of my tenons in a mortise just to get an additional gluing surface. I will glue the tops of my dovetails. Great videos all around, thanks!
The test didn't show that end grain glue joints are stronger so you really don't need to find more end grain involvement....
I watched it last week and his outcomes were very surprising. I posted it to my local FB woodworking group. It surprised a lot of woodworkers. I think it was very well done.
Thanks for validating my hard work.
was the surface area equalized in the glue only and mortise/spline joint? Much of the differences could be from more surface area involved in the mortise and spline joints.
Isn't the simple way of saying the whole thing something like: if you need a wider panel, you can glue side to side, and be safe in the knowledge that even if the panel is glued up, the lignin will fail (well) before the glue joint, so nobody will ever complain about the fact the panel is a glue up -- any failure will be in the wood, not in the joint. On the other hand, if you need a longer panel and glue endgrain to endgrain, that *will be failure point* -- because you have introduced an obvious failure point, should a failure actually occur everyone will say you are an idiot. Also, it is easier to find longer panels, which avoid the issue of endgrain to engrain glueups, than wider panels, so glueing side to side makes sense practically and structurally, whereas glueing endgrains does not.
Well said, I was trying to think of times when you can't realistically just find a board that is long enough for whatever you're building and the only situation I could think of was boat building.
@@WobblycogsUk that and timber framing. And still on those cases they use massive scarph joints, which again take advantage of the resistance of wood and not the glue.
I recently discovered that one of the 4 by 4 I bought in the woodshop was glued somewhere in the middle. However they cut two sawtooth patterns in both ends and glued them together.. Dunno if it as strong as the whole 4 by 4. Didn't notice any difference and prolly wouldn't have noticed if not for grain color.
Yep - and they would be right in saying it is your fault - cause it would be your fault.
glueing endgrain to endgrain is a weak JOINT, just it is not the case that the glue in it is magically weaker but the joint it self is weak as it does not have the continuous fibres that make the wood strong along its grain.
scarf joint or lap joint - those should have been compared to the simpel endgrain glueup - that is what the saying about endgrain-joints is all about.
Been doing middle age European style timber framing for years now, I have never thought to just glue two pieces together endgrain to endgrain because here all of this (side to side lignin will fail, end to end glue bond will fail and along the grain will take the most to break the wood- that is why joinery came to be) was a part of 6th grade woodworking class. To be honest in classical timber framing one really doesn´t need glue or metal fasteners as wooden pegs and the right types of joints with no glue in them will do all the hard labour of keeping the structure solid. Sorry if my terms are off, English is not my first language.
I didn't see the video but saw the people making the assumptions you did. Some said it tongue in cheek and I think a few believe it. As you point out here, it isn't data for wood working.
Here is something you know, but maybe helpful to a novice.
Wood is made of fibers. If you break it, it splinters everywhere. The fibers keep it together until they separate from each other and break at different places. To break a board, you have to separate the fibers.
In martial arts, they break boards by exploiting the weakest angle to the fibers.
When you cut a board, you weaken the fibers on the edges. When you glue it together, the glue bonds the already broken fibers, so when you break the joint, the fibers near the joint, not far away, break. The glue may not be as strong as the center of the board, but it is stronger than the edges.
End to end tinsel strength, I can see that working. You have the ends of thousands of fibers gripping the glue and no leverage. The collective fibers are giving it strength. But, like you proved with your knee, the board didn't break. The glue broke because you separated the outer edges first and then all the rest gave way.
I broke a bunch of but to edge glue joints today from older builds I abandoned or replaced. I used leverage and the glue almost always kept the fibers of the edge board. In over a dozen joints, only the edge of the board where it met the glue gave way.
BTW, just to be clear, the joints were pocket hole joints that were glued. I removed the screws and reallocated the wood to my current project.
I'm glad you made the video because it's good to see what the talk is all about.
Thank you.
So glad you made this video. After watching Patrick's video, I could already envision the furniture/projects for sale on FB with end grain joints.
Great video to add on to Patrick’s findings. While I complete agree with what you had to say wouldn’t the thickness and width of the wood being used have an effect on the strength of the joint? I would be curious to see the results of the strength of the glue joint between two boards being glued in a cross form or a even in a half lap type joint. Again great video Marc.
To me the key point to take from both Patrick's video and this one is that lignin isn't all that strong. Don't rely on that if you are going to have a lot of demands put on a joint.
The secondary point is that the glue is stronger than lignin. So properly done, an end-grain to end-grain joint will be pretty strong compared to what we were taught to expect.
Now I want to see someone testing scarf joints. . . There's probably a video on that as well.
as a structural engineer, the next step to test this would be to do straight tensile tests... as the other video pointed out in the end to side tests, the stiffness contributed a lot. therefore, tensile cares very little about stiffness, at least compared to bending (planer sections remain plane assumption).
Though we'd probably still find the same thing - it was a statically determinant beam after all. just want to see a test trying to eliminate stiffness as a factor and truly test the glue bonding to the different types.
Love the shirt! I really like her videos and creations. But back to the subject at hand, I think you did a very good response and reaction video, showing how even though he isn’t wrong, we still strive for stronger joints than what only glue can provide.