I never realized people used biscuits like that to add strength, I've only ever used them to help with alignment when gluing long edges together, as the depth from the top surface with the biscuit cutter can be easily set.
This is how my dad taught me. We mainly used it for alignment on desks, or similar furniture, as there are many long pieces that need to be glued together, and biscuits are the easiest method. Not for structural integrity, but for lining up, speed, and enough structural integrity to make the rest of the piece.
Another vote for biscuits as alignment tools only. Dominoes are a different story, from what I’ve seen. But I’d rather chop mortise and tenons than shell out $1000 for a Domino :)
An an amateur I had assumed years ago this was the entire point of them. Using poorly cut (certainly not planed) lumber I was able to glue up a table top relying on the biscuits to help keep it aligned. They prevent the clamped surfaces from sliding across one another. There are other techniques to deal with that but my 2x8” table thrown together with bits from Home Depot and clamped on bendy pipe clamps came out surprisingly well and looks nice years later.
@@patty109109 Honestly biscuit slots are so sloppy I don't know if they even serve that purpose well. In my experience their primary advantage is efficiency.
Interestingly, we found similar results when adding reinforcing fibers to concrete. I work for the CA DOT testing cement and concrete. A number of fiber reinforcing products have come around over the years. I'm talking like 1/2" to 2" fibers, not reinforcing steel. There have been plastic fibers, glass fibers, and metal fibers. Seems like it would make the concrete stronger, right? Well, it never seems to increase the maximum strength, because the fibers don't really get involved until there are significant displacements inside the concrete, and that doesn't happen until after the concrete breaks at maximum strength. BUT, what the fibers do well is holding concrete together after it breaks, much like the residual strength Patrick found here.
@@sawdustadikt979 The last time I took a concrete sample cylinder was, oh, 1971, but the dynamics of concrete construction (and by "construction" I mean the combination of compressive and tension elements, which include cement, aggregate and reinforcing) haven't changed (and "dynamics" is the key word, as contrasted by "static" analysis). There are several things going on inside a concrete slab/beam/etc: the rebar provides tension strength, in part by depending on the compressive strength of cement caught between the rebar ribs (which anchors the rebar to the cement); another dynamic is the adhesion of the cement to the aggregates (technically, concrete is composed of cement and aggregates--rock and sand). Any section of concrete construction placed under stress turns out to be under both compressive and tension stress. The shape of the concrete and the direction/magnitude of the forces determine the amount and location of those two stresses. The biggest failure mode for concrete is the loss of adhesion between the rebar (the tension element) and the cement; very frequently the loss of adhesion is caused by water penetrating the cement/concrete (rock and cement will absorb water and capillary it to the rebar), which turns the (iron) surface of the rebar into rust, which just flakes apart. Salt water, as found in coastal areas with high water tables, rusts iron very quickly. Generally, concrete is qualified by its compressive strength, e.g., "3000 psi concrete".
Hi Patrick, I'm not a great woodworker, but I do know a little about stress and strain. the way you tested the biscuit joint creates bending through the piece, which means axial tension is generated at the top surface, and compression at the bottom surface. because the biscuit is in the middle of the piece, at the neutral axis, it actually doesn't see much of the bending stress, so it's not contributing much to the strength of the joint. aside from helping with positioning the pieces, what I'd expect the biscuit to help with is transfering the stress from one side of the glued joint to the other when you're putting the bending across the long aide of the wood (similar to the prying apart of the joint which you showed in the images at time 7:30 in your video). I'm not sure how great the effect is, but if more of the force is transfered into the fibres in the biscuit, then from there into the fibres on the other side, it would probably do more for increasing the bending strength of the joint... the top half of the biscuit would be transfering axial. hmmm. didn't explain that well but the summary is that for your tests, the piece was bending around the biscuit, with the biscuit being in the neutral plane for that load orientation
A second test based on the above with 2 biscuits could give an answer to the above. I do love the testing and results and myths busted! Thanks so much Patrick for this series :-)
Shaun, I found myself asking: how would the biscuit affect tensile strength? My initial thinking was that the additional glue area created by the biscuit doesn't change the glue area along the plane of the failure, but that is pretty shaky. I think your analysis is correct in terms of why the peak force to failure in bending doesn't change. But since the mechanics of wood and glue are so complicated, there is an additional set of head scratching questions. Since the failure here is brittle, boundary conditions are critical: say we move the biscuit as far from the neutral axis into the peak tensile stress, that would effectively mean just slapping it on top and across the joint! That changes the mechanics entirely, since it would be far more difficult for a crack to initiate with the long wood fibers in tension. Where would it fail then? So, while it could be interesting to see the test repeated with a biscuit at different distances from the neutral axis, I think it's maybe more fruitful to think about the broader class of splines, inlays and the like. Each will change fracture mechanics under different loading conditions with their own benefits and drawbacks. It would be fascinating to see different methods compared through tests under a variety of loading conditions: if well done, that could add some real engineering foundation to woodworking practice.
@@jiijijjijiijiij Except that this series is about glue and not about the geometry of the joint. He specifically says that these results only apply to stress as generated by “my rig.” He shows at the end a number of force vectors that are typical but not tested.
@@blacksmith67 you’re right, but it is unfortunate that he went to all this work but then applies the force at a direction that a miter joint like on a picture frame would never have
I do Agrar, tha a biscuit in the neutral axis should not improve the bending strength. In practice, joints with biscuits are stressed often perpendicular to the biscuit plane. That could be a test worth to do. Or biscuits are loaded with a force parallel to their plane, eg. if put in two rows into a thicker board and away from the neutral axis. So I will continue using hundreds of biscuits also in future, and I am confident they will hold things together. Btw: Since I live near Basel where the original Lamello was invented, my English might be only semi-correct
I think the main reason why the biscuits are made in an angle is because you don't know what direction you are going to need the biscuit. Most cuts in woodworking are rips and crosscuts. So having the biscuit in a diagonal covers both options
I use biscuits as an easy helper to keep wood aligned, especially on heavy long boards. Paired with pinch dog clamps, biscuits really help when your panel clamps just aren’t long enough or you just don’t want to use them. I’ve never viewed biscuits as a reinforcement but rather as an aid for aligning.
Yes. I use them solely for alignment purposes. Especially when doing long edge grain panels. They aren't perfect but reduce the alignment error greatly and eliminates having to deal with slipping when tightening clamps. The little time it takes to cut the slots and install the biscuits totally makes up for the fuss of aligning panels and the amount of sanding afterwards to smooth out the joints.
I thought this was what everyone believed until I started seeing UA-cam videos claiming they don't "add strength." It was always just to keep panels aligned. I blame Festool.
Exactly, you don't worry about your table holding up an elephant, you worry about board warpage and glue failure over time. Plus mitre joints don't need to carry weight, they need to withstand being pulled apart length ways.
More like it is silly quasi science. First video was pretty dumb and here he tests biscuit joint strength in very silly way, in end grain joint which is weak and unusable in butt joint. Only way to joint end grain strongly and reliably is to increase glue area - in practice this means avoiding end grain joint. In construction lumber never ever butt joint would be accepted, it is so weak . His conclusions are just stupid
I just watched your glue series and was impressed by the experimental setup that allowed head to head comparisons among the different woods and joint orientations. Avoiding differential calculus in favor of a solid "4X stronger" evaluation was a great choice enabled the design of your samples. Very informative, great presentation, fantastic attention to detail.
Unfortunately you’ve tested these in bending with the biscuits at/near the neutral axis (middle) where stresses would be near zero until after failure of the top/bottom of the specimen. Failure of the top/bottom would largely not be affected by the biscuit at the neutral axis. You’ve indicated the limitation of your force application, but you didn’t limit your conclusions to reflect them. To use your bending setup effectively, maybe consider putting a pair of biscuits so they are located away from the centre of the wood specimen to see if they add strength when actually being stressed before failure. It is nice to see a more deliberate effort to pull apart some of these glue myths, keep it up! 👍🏻
I have to agree with Pat. Your test only covers a case where the biscuit doesn't add significant strength. The other cases that you leave for others to investigate would probably result in a very different conclusion. I would argue that in the real world, a good proportion of the stresses on miter joints probably line up with the "other" cases, and as a result, the average strength of a miter joint would increase with the addition of a biscuit (although I'll admit that a splice is still superiors when it comes to adding strength) . Your notes about the grain orientation of the wood in the biscuit is interesting. I wonder why the manufacturer does it this way. This brings up an interesting question: What would happen to the strength of a joint if the grain in the biscuit was oriented properly? What if a hard(er) wood was used to make the biscuit? The fact that it tore itself apart during your test means that there is room for improvement in the biscuit... Thanks for the video, still very interesting!
@@NugTrace I'm tempted to make up a handful of specimens and run them into the lab where I have the capacity to test them in uniaxial tension. Might be interesting... sadly there are so many other things I'm supposed to be doing!! 😁
Agreed. But I also wish he'd modify he's load testing to isolate shear and tension strengths as separate tests. His set up induces a high shear to bending moment ratio that is atypical of actual applications. And a biscuit won't contribute to any additional shear strength since the extra glue area is outside the shear failure planes. I'm not sure how much extra strength a biscuit actually would add, but this test is virtually designed to render the biscuit useless.
@@aaroncecil7509 His short 3-point bending setup indeed, as you noted, make for coincident shear/normal stresses at the point of the joint. If COVID allows, I hope to run a handful of specimens through pure axial tension tests in the new year to see how they will do. I don’t know how they will compare, but at least we will see the biscuit joint being pulled apart. Should make for a fun 4 or 5 min video.
My understanding was always that biscuits add no real strength, rather are used for part alignment only. Love to see you test dowels and dominos as a comparison.
I like dowels. Mainly because it's a cheap system and easier than tenons. I would predict dowels, dominoes, and tenons to have similar abilities because its basically all the same concept
I would use biscuits not for alignment during assembly, but an insurance policy against boards that want to warp in the future, especially trim pieces.
One point to consider: the biscuit was inserted to the neutral axis formed by the bending moment. I think this location plays a bigger role then it's mechanical properties or the gule.
@@JohnDotBomb These tests desperately need to isolate shear and tension as different tests. His load testing creates stresses that would not ever be induced in actual joinery applications. This one for instance induces a high shear to bending ratio (and loaded in the typically nonstressed axis to boot). All the extra glue surface area is rendered useless because it is outside the shear failure plane, and as you said it's extra tensile properties have to impact because it is in the bending neutral axis. Then after the thing has failed he finally puts it in tension when he tries to pull it apart and is shocked it is actually suddenly doing something.
As others have said, alignment is the issue when I pull out the biscuit joiner. Almost always in situations where I have side grain to side grain - not end-to-end grain. The myth seems to be that people use biscuits for strength since most comments state they have never relied on them for strength. The series does help us all in that it is consistent in method and either reaffirms what we suspect or blows up the folklore of others. Great series with a lot of lessons.
mitre joints are not designed to resist or withstand the direction of the forces applied in these tests. they're meant to go around corners, so the forces they must resist are those that try to deform them in the (x,y) axes by closing or opening the angle at the joint. this press test applied force only in the z axis, which is not what mitres are designed for. biscuits and splines strengthen the joints in the (x,y) axes, which you proved when you could not pull the pieces apart by hand. i bet if you tested mitres with and without the reinforcement of biscuits you'll find a huge difference.
Thanks for articulating my thoughts exactly. I can't think of a situation in woodworking where the stress on a joint like this would be applied from the side.
@@bekanav 100%. It's a shame how many people think this is super helpful information. The information gained from these has no application to what woodworkers are doing. If anyone is making mitres the way he is in these videos they need to rethink what they're doing.
@@fabianherrera5427 Yeah, he mentioned it. But he still went on as if his testing was valid. The methodology was flawed, non-representative of real-world conditions. So his conclusions are entirely irrelevant. An accurate test has to simulate a real use case.
Your 3 point bending test illustrates a simple engineering concept for beams. Loaded beams have one side in compression and one side in tension. In between is a thing called the neutral axis, where the stress is zero. In your case, because the biscuit is located in the middle (which happens to be the neutral axis for this cross section), the bending does not even put stress on the biscuit until after the butt glue joint has failed.
I felt like something was off in this test, and I think this hits the nail on the head. The advantages of the biscuit are not well represented in this test. I would have liked to see a tensile test of the miter joint with and without a biscuit to see a comparison. But since the grain direction of these biscuits are 45 degrees, they may not be as helpful as I thought.
This is a good point, but the neutral axis is an infinitesimally small line. Biscuits are not that. It is possible that the biscuits break even with a small fraction of the stress that's applied on the outermost surfaces before those break.
I was going to mention this. Most of the strength of any structure for most types of forces comes from the outer surfaces. If you're using it close to the limits, you could epoxy much strong dowels deeper into a joint, like steel rods.
Thank you for the knowledge, care and effort you poured into this gift. Your videos are required viewing from now on. I appreciate that you point out lignin's comparative weakness as an adhesive. PVA is a modern marvel. I especially admire, and appreciate how you focused the content so that each video could be limited to ten minutes, the optimal viewing time for retention. Of course, the points made raise further questions, which underscores their value. I can't wait for the next one.
Thanks for sharing your very thorough and logical research. I'm new to woodworking and as a retired layabout have lots of time to watch these myths being put forward in various ways. Your three simple videos are incredibly informative and most reassuring. I'm not an engineer, just enjoying the beauty of wood and what you can do with it, these videos are brilliant.
Great videos Patrick. I know some others have suggested it, but it is "common knowledge" that domino joinery is just as strong as a traditional mortise and tenon. I've always doubted this, but I would love to see a test done. Keep the videos coming! Woodworkers have been doing it right for millennia, but technology has changed and anecdotal truisms can be tested!
I watched a video about what makes the material strong when forces are applied like this. One surface is under compression, the other is under tension. Most of the material in the middle does not contribute too much. So the only way to increase strength is to put something on the surface. even thin veneer will do better than biscuit. However ff you are making a frame a spline will add to the strength, because as you said there are other directions that forces can be applied from.
Mr. Sullivan: I very well done video....the filming, voice over and graphic work is top notch. Thank you. I would only say one thing though about the content and that is, we don't use biscuits for strength, we use them for alignment. If I'm trimming a window and using 1/4 stock or mdf flat stock or whatever I may use a biscuit so that the faces align.....not because of strength. The glue is strong enough on it's own. I trimmed a house once and after finishing the lady decided she wanted shiplap everywhere so a guy came along and took down the trim around most of the windows and doors. They were coming off like picture frames ...in one piece and they were only glued (at the miters). Glue is plenty strong on its own.
Hello Patrick Sullivan. As a mechanical engineer I can tell you why a biscuit wouldn't add strength in your test, even if the biscuit was made of something much stronger. Whenever a given object is experiencing bending, the actual stress in the material is NOT uniform across the thickness of the object. It is the highest the further away from the center (oversimplified) you are. And as such, a biscuit put in the center of the boards won't ever get to help reinforce the joint when it comes to bending. Perhaps a test with a board thick enough to have two biscuits in the height plane could prove that biscuit either add or subtract strength. The biscuits would need to be close to either surface instead of bundled together in the center. So perhaps a 20mm thick piece with either biscuit being only 4mm below the surface would prove to be stronger or weaker than without.
I think this test shows the limitations of your set up. It's worth doing these tests again but instead testing the pure tension strength of the glue joints by seeing how much force it takes to pull the joint apart
Haha, you had already addressed my thoughts concerning the previous video before I even thought of it. Great video! Very useful and informative, as the previous ones. Thanks!
You nailed it when you state the area vs the direction of the force applied equal the strength of the joint. With a biscuit, the area of glue is actually DECREASED when you are talking about the direction of the force applied. This means the amount of strength gained is about equal to the amount lost to the lack of glued surface in relationship to the direction of the force applied.
Patrick, Love it. I would also love to see results in pure tension and shear too, if you could do it. The bending moment butt joint to failure is actually a combination to the two. Envision the top of the joint is in tension and the bottom is in compression over the fulcrum. Also, try smearing some glue as deeply as you can by rubbing it into the end grain and then adding a bit more glue. I've done that on miters in red oak and it seems to be even better. Did kitchen cabinet doors like that (no splines or biscuits) 30+ years ago and they still are together. Thx for the testing!
It was my understanding that biscuits were for alignment and a bit of strength for butt joints (ie. cabinet rail styles, frames etc) or slab glue ups. Very interesting info, thanks! I made a solid oak table, used biscuits to help align the oak planks for the top, it has never cracked or separated along the joints. Later I needed a quick T.V. stand, used biscuits as shelf supports, glued and clamped. let dry. That stand held an old heavy 27 inch tube T.V. for a few years before T.V. was replaced and is still holding together.
The purpose of the biscuit is for alignment during glue-up and to hold the project together even if the glue joint fails. Also, biscuits should be used with a water based glue to work properly. Biscuits are compressed and the water is absorbed by the biscuit making them expand to provide a tight fit and proper alignment. I've glued up biscuit joined pieced and then needed to pull them apart several seconds later because I goofed up and they would not come apart because the biscuits swelled up and held everything in place. I switched to Titebond III and noticed that my parts still had a lot of movement to them even several minutes after being glued. The reason being that itebond III is a polymer based glue and if it does contain any water, it's not enough to swell biscuits.
Helpful! I've been resisting buying a biscuit joiner, although I've used them before. I know they have their place, but I think there are better alternatives. Thanks for this video!
Biscuits are great for aligning pieces to eliminate "lipping" and making it easier to clamp mitered frames together as they minimize slippage. So they do have a use, you just have to understand that when deciding whether or not to use them. Same with floating tenons popularized with the festool domino
As someone else pointed out, the biscuit was inserted at the neutral axis. Before the test was conducted I expected the result would show no difference from previous tests. In a bending load the max stress is at points furthest from the neutral axis, which means that peak load depends on glue strength, and the biscuit's contribution is negligible. A complete set of loading conditions need to be tested before meaningful strength conclusions can be made. However, even with that information, I doubt whether any wood joinery design would be altered.
As always biscuits are tested in a manner that is completely unrelated to their design purpose. I think the general public really misunderstands this tool.
First test was a bending test along the minor (short) axis. The peak stresses would be at the edges with only the glue bond. These stresses would be the same with or without a biscuit. A tension test or bending along the major (long) axis may be more telling for just the biscuit strength itself, but still won't help the table strength with end grain connections. Conclusions about fit-up are good case. I know people that just use dowels over the biscuits or dominos. Be interested to see if dowels are any different than biscuits.
Also the cabinet shop I worked in as a young man we would use 23 gauge headless pin nails four on each side to bind the biscuit and joined wood four pins per side.
Pat H has identified the exact weakness of your testing method. Think of a steel I beam. The top and bottom flanges are wide in order to put most of the material (strength) where most of the stress is when a bending load is applied. A very small thickness of metal is in the center of the beam (web) the stress from bending is lowest along this line (neutral axis). There is no location in the joint where the biscuit contributes less strength to a joint in bending. This kind of testing is important so please keep up the good worl,
Going back to old Norm Abram material (who obviously deserves a lot of props for making this possible for so many of us to discover, so bless him for that) from the late 1980s, there was definitely a time when we thought biscuit joinery was almost magical. It’s amazing how eager we are to think that of new things. As usual, these tests in the end are most of all a reminder to think through what our assumptions are and ask if they actually make sense in the context of what we’re doing.
I suspect it's not so much a matter of "glue area doesn't matter" as "glue area where the stress concentrates does matter". It's the difference between static and dynamic analysis (ouch, using the Wayback machine to 1970!). Any force applied becomes tension in one area and compression in another, and the exact location of the critical stress can be deceiving. The location also moves as the stressed structure deflects under load. The 45 orientation of the biscuit grain is probably a clue (manufacturers didn't arrive at that orientation by accident). Biscuits might contribute more to joint resistance when the joint is loaded on a different axis than the simple end grain reinforcing axis. However, at least it's good to know biscuits don't reduce the joint strength. Excellent videos Mr. Sullivan. And to your profession: "thanks for all the fish". (I'm a forensic engineer!)
Been waiting for this one! My guess is that the glue joints of the biscuit to the slot don't fail - the glue area that counts is the butt to butt joint is actually lower, and the biscuit itself cracking is proof that it's adding about the same resistance to the glue lost, not any additional. i think that's my takeaway from your testing too... and i think the point about facilitating assembly and getting good alignment between parts is right on the money. either way another great, data driven video that had my riveted
I bought into the biscuit joinery fad back in the 80's and spent a hefty amount on a Lamello, it took about 6 months for me to realize that it didn't improve my work or provide any benefit to what I was doing so I sold it and never looked back.
Hang on. You tested bending strength about the centre of the joint. Had you tested pull out tensile force then biscuits would have reinforced the joint. Also biscuit material/ strength varies. I’ve used some quality biscuits which do most certainly add strength. A few too many simplifications I think in this video.
I have always thought of biscuits as a way to “locate” joints. Since I have learned the technique of dry-fitting joints with blind pins, I no longer use biscuits to avoid slippage.
Thank You for your excellent video I seldom use my biscuit joiner except for difficult joints never for strength. Small tip for biscuits keep them in a sealed bag or container add descant pouches for added protection from moisture.
Biscuit/plate joinery in many shops is mainly to speed production and glue up, not necessarily for strength. I've always thought of biscuits as strong though, generally.
Another excellent review of a wood glue joint consideration and examination. I which you would do the same test with 2 or 3 dowels in the joint vs biscuits since you can do the testing and I cant. Hope you do to see the results on that set up. I have enjoyed your video and these studies. I have never understood biscuits and the reasoning behind them. 5 thumbs up from me Patrick! Columbus Michigan
Patrick, thanks for this clear and understandable video series. I hope your viewers' take-away is that woodworking joints should be designed to rely on the interlocking structure of the joint for their strength against the forces at play during the object's intended use. Glue is intended to prevent the joint from slipping apart in the reverse direction of assembly, which should be in the direction LEAST likely to have excess force applied during use.
I use biscuits to keep things aligned in clue joints that I suspect will break. For example on a chessboard along the endgrain. I know from building many chessboards in the past that it will have between ⅛ to¼ inch in movement so there's going to be an endgrain joint somewhere that will fail. The biscuit is there to keep that joint aligned and intact.
Really love what you’re doing, fantastic data. I was always taught to keep a open mind and never stop learning. Thank You for this series looking forward to seeing more information!!
Patrick, Your series on joint strength has rattled my long held beliefs. Very interesting and makes me wonder about the double tenon joint strength. Festool is a common system that I’ve believed to be superior to the biscuit type joints. Can you please do a series on their strength in one and two row configurations? Thank you for your exceptional videos and please keep exploring myths/truths about best construction techniques and possibly glue types and properties. Cheers, Will
In think leverage is everything and limiting it with grain orientation is the most important factor. No matter how strong the joint you can break it easier if leverage is available to the joint because of the finished project allows it.
My favorite application for biscuits is attaching a face frame to a plywood carcass; biscuits have pretty much entirely replaced brad nails for me in this application for several reasons, including biscuits are stronger than brads while the glue is wet, and there are no nail holes on the finished exposed surface of the project. Here's one thing I wonder: would adding a biscuit (or any other spline) increase the strength of a miter joint not by adding its own strength, but by helping align the joint and thus forming a better glue bond? Probably not in your test rig where a "miter" joint is mechanically a butt joint but in a real "two boards with 45 degree ends meeting at a 90 degree angle" situation. Those joints are difficult to assemble and clamp, so adding a biscuit or spline to constrain the two boards together may allow for a superior miter joint. It's one of those things where the "I made some joints and then used a bottle jack and a bathroom scale to break them" tests kinda fail; I think this methodology eliminates variables that actually matter in practice.
If you want to treat a biscuit like a piece of rebar then you would add the biscuit as close to the outside edge as possible. Adding strength as close to the center as possible does little to prevent bending forces.
This is actually an important takeaway 7:16. This test, like many others have said, did not test the type of failure that when used in a miter joint. The test and its conclusions are valid though in the case of panel gluing, due to the biscuit being placed in the neutral axis of bending.
Ha ha, I knew or lets say felt that biscuit joints doesn't do any much help to strength :) so I never bought one and I always thought they look stupid... But it seems you have more opposite comments for this video compared to the first two videos and I guess that is because many new wood workers have been convinced to buy a biscuit machine and using them since a time and don't want to accept that that machine is not much use except for the aligning purpose. But for my view, If you cut the 4 pieces identical to each other, there will be no problem with alignment and you can also always hand plane them a little and there you go, you have a perfect miter... Thank you very much for the video 👏👏👏👏
In the shop where I work, biscuits are used primarily for alignment. We build drawer boxes out of melamine by placing a biscuit or two in the corners of the drawer box. These joints also get glue on the spline joint as well as melamine glue on the butt joint itself. Finally, the joints are stapled together. We never have any failures with these joints. I have taken some of these drawer boxes apart and they always break apart with the particle board still glued to the biscuit. In this case the biscuit is stronger than the particle board around it.
I was given a large divided cabinet, melamine covered plywood. all the dividers and shelves were biscuited, I had to cut along the shelf gap to get apart. the glue hadn't stuck to the melamine but the biscuits sure put up a fight. When I did get some shelves to come out, the plywood/ glue failed around the biscuit or some just broke. Was crazy how hard it was to take apart.
In the direction that you were testing the samples, the pressure on the joint would tend to split the end-grain joint on the OUTER surface first, and then very rapidly the split would tend to progressively work it's way into the biscuit (this is milliseconds fast). Then the moments acting on the straight grain of the biscuit would tend to split the fibres apart, as they rotate, to try and reach the direction of the pulling force acting on them, thus tearing the biscuit and the joint apart. So it seems likely to me that the best thing to do, is not to put ONE biscuit on the centreline of the thickness, but to put TWO biscuits in, just under the surface of the back and front faces. That way the biscuit will act like plywood just below a thin veneer. I wonder if you tried that, the results would be different???
Excellent work once again Patrick!! I had watched part 1 and 2 and wanted to see the spline effects. With my Mechanical Engineering background, I have to totally agree with the comments about how the force is being applied, i.e., along the neutral axis, and that is not the practical application. As most everyone has stated, the biscuits, dowels, and dominos are all used for alignment while gluing. It would be interesting to see what would happen if there were (2) biscuits evenly separated away from the neutral axis, but I don't think your test sample it large enough to allow for that? However, could you turn all cases of your test pieces 90 Degrees to simulate more of the breaking force that would pry open the joints over a longer distance? I truly enjoy and appreciate your work. It is all accurate in the approach you have taken, but like prototypes, we learn new setups as we go and come up with ultimately a better product. Thank you very much for providing data versus just opinions, which everyone else can easily do. LOL
I made a mortising jig for my router. It uses a bushing that rides in a slot at the top. I based it on Phil Morley's jig, but I made it so I can cut pieces of all angles (though the top has to be flat up against the bottom of the base), so my jig uses t-tracks and blocks with clamps on them, so I can do very narrow pieces (down to 1/4" mortises on 1/2" pieces of wood) or and I can do very wide mortises, and/or cut into pieces that are very wide, so the mortise can be set back far from the edge. The top moves about 6", so I have a wide range of choices... I then custom make loose tenons to fit into those mortises. This makes me want to test the breaking strength. I like the loose mortise and tenons because the mortises are fast and easy to cut, especially now that I've designed and built my jig, and the tenons are also very easy, and can be cut with repeatable accuracy. I feel they are better than dowels because they should resist any racking/twisting more than dowels, and part of me wonders if they may be structurally better than hand (or machine) cut mortise and tenons that are not the loose kind. They certainly provide much more structural stability than biscuits. I built a gate this past summer, when I replaced my fence. The gate is 9'6" wide, and in the middle is a W, with those four straight lines of the W each being a piece of lumber, with a loose tenon attaching them together, and I was able to lift all four 2x4's with the tenons in place, and the W held its shape, and that was before I had having put any glue into them, and before I had put the W into the frame of the gate. In any case, I appreciate your expertise - you clearly have more experience with this than I, and it is interesting to think outside the box, and you did, literally, break the mold, so thank you for that!
In my opinion, when seeking to add strength first understand where (the direction) your joint is expected to experience force, then look to how you can strengthen.
very good work. i feel i need to wait the end of these series of videos de make conclusions, once all de preparation for each video is so laborious and there for takes time. also i see you are building a train of thought and are wise enough to make the choice for each subject. i am very pleasy for your work. thank you a lot.
However, the LINEAR strength of biscuits adds to the linear tensile strength.The glue joint has 2 axis lines, and the pull strength of this joint is actually improved. Then again biscuits also align the joints so glueup is easier and faster.
I wonder if maybe the reason the biscuits themselves were cut on a bias was solely to decrease breakage and waste in the manufacturing process. If they cut them with the widest part of the biscuit as end grain (the theoretically strongest way for them to cut them as far as the joint is concerned) I'd bet that alot of them would break in half just tumbling around conveyors etc. If the cut them the other way they might last longer in the factory but would be actually useless for the end user. The 45 deg cut is a compromise.
Oh man, oh man, another episode! Can’t wait for the juicy debate to follow. These videos have shaken up the UA-cam woodworking community in unexpected ways like publishing the special theory of relativity or early behavioral economics!
Using biscuits to add more gluing surface to a joint is like cutting your boards in half length wise to add more gluing surface. Technically it will add surface, but it doesn't add any strength to the joint. I've always viewed Biscuits as a device to help align boards together. If you need additional strength, a half lap or tongue and groove type of joint will probably be better.
Thanks to a biscuit-joiner esque tool from Grizzly, I've switched entirely to dowels with most of the efficiency of biscuits but greater strength (as far as I can tell so far
I'd imagine that a biscuit would add significant tensile strength to a but or miter joint, and ultimately most forces applied to a miter joint would stress it along it's tensile axis rather than its sheer axis. Maybe there's a way to devise a test contraption that will pull apart a miter joint perpendicular to the glue line?
The results surprised me. I expected a small increase in the joint strength with a biscuit, but I still expected the biscuit joint to be weak. Two of the things going on here are: 1. The biscuits should be installed on the outer edges if additional. That would be ugly, so I'm not suggesting that this be done. 2. The direction of the fibers of the piece used to strengthen the joint need to be across the cut. The problem with a butt joint is that there are no wood fibers that cross the joint. Biscuits would help more if the grain direction of the biscuit were changed to be parallel with the grain direction of the wood. The bottom line is that almost any normal wood working technique would strengthen a butt joint more than a biscuit. Biscuits are bad because the are thin so they concentrate the strength they add almost entirely in the center of the joint. Biscuits are bad because they grain direction is wrong for a butt joint. And biscuits are bad because they add less glue area to the joint than most other wood working techniques would.
I never realized people used biscuits like that to add strength, I've only ever used them to help with alignment when gluing long edges together, as the depth from the top surface with the biscuit cutter can be easily set.
This is how my dad taught me. We mainly used it for alignment on desks, or similar furniture, as there are many long pieces that need to be glued together, and biscuits are the easiest method. Not for structural integrity, but for lining up, speed, and enough structural integrity to make the rest of the piece.
Another vote for biscuits as alignment tools only.
Dominoes are a different story, from what I’ve seen. But I’d rather chop mortise and tenons than shell out $1000 for a Domino :)
Over 40 years of woodworking under my belt and I've had little use for biscuits thus far.
An an amateur I had assumed years ago this was the entire point of them. Using poorly cut (certainly not planed) lumber I was able to glue up a table top relying on the biscuits to help keep it aligned. They prevent the clamped surfaces from sliding across one another.
There are other techniques to deal with that but my 2x8” table thrown together with bits from Home Depot and clamped on bendy pipe clamps came out surprisingly well and looks nice years later.
@@patty109109 Honestly biscuit slots are so sloppy I don't know if they even serve that purpose well. In my experience their primary advantage is efficiency.
Interestingly, we found similar results when adding reinforcing fibers to concrete. I work for the CA DOT testing cement and concrete. A number of fiber reinforcing products have come around over the years. I'm talking like 1/2" to 2" fibers, not reinforcing steel. There have been plastic fibers, glass fibers, and metal fibers. Seems like it would make the concrete stronger, right? Well, it never seems to increase the maximum strength, because the fibers don't really get involved until there are significant displacements inside the concrete, and that doesn't happen until after the concrete breaks at maximum strength. BUT, what the fibers do well is holding concrete together after it breaks, much like the residual strength Patrick found here.
I thought the main use was to limit breaking of the concrete while it dries. Is that not a thing?
Am I getting it right that steel , base and the right mix is the formula for resistance to breaking?
Am I getting it right that steel , base and the right mix is the formula for resistance to breaking?
That is fascinating! Thank you for sharing!
@@sawdustadikt979 The last time I took a concrete sample cylinder was, oh, 1971, but the dynamics of concrete construction (and by "construction" I mean the combination of compressive and tension elements, which include cement, aggregate and reinforcing) haven't changed (and "dynamics" is the key word, as contrasted by "static" analysis). There are several things going on inside a concrete slab/beam/etc: the rebar provides tension strength, in part by depending on the compressive strength of cement caught between the rebar ribs (which anchors the rebar to the cement); another dynamic is the adhesion of the cement to the aggregates (technically, concrete is composed of cement and aggregates--rock and sand). Any section of concrete construction placed under stress turns out to be under both compressive and tension stress. The shape of the concrete and the direction/magnitude of the forces determine the amount and location of those two stresses. The biggest failure mode for concrete is the loss of adhesion between the rebar (the tension element) and the cement; very frequently the loss of adhesion is caused by water penetrating the cement/concrete (rock and cement will absorb water and capillary it to the rebar), which turns the (iron) surface of the rebar into rust, which just flakes apart. Salt water, as found in coastal areas with high water tables, rusts iron very quickly. Generally, concrete is qualified by its compressive strength, e.g., "3000 psi concrete".
Thanks Michael. Like Switch & Lever below, I always used biscuits for alignment and with my coffee. Nice work.
Hi Patrick, I'm not a great woodworker, but I do know a little about stress and strain. the way you tested the biscuit joint creates bending through the piece, which means axial tension is generated at the top surface, and compression at the bottom surface. because the biscuit is in the middle of the piece, at the neutral axis, it actually doesn't see much of the bending stress, so it's not contributing much to the strength of the joint.
aside from helping with positioning the pieces, what I'd expect the biscuit to help with is transfering the stress from one side of the glued joint to the other when you're putting the bending across the long aide of the wood (similar to the prying apart of the joint which you showed in the images at time 7:30 in your video). I'm not sure how great the effect is, but if more of the force is transfered into the fibres in the biscuit, then from there into the fibres on the other side, it would probably do more for increasing the bending strength of the joint... the top half of the biscuit would be transfering axial.
hmmm. didn't explain that well but the summary is that for your tests, the piece was bending around the biscuit, with the biscuit being in the neutral plane for that load orientation
A second test based on the above with 2 biscuits could give an answer to the above. I do love the testing and results and myths busted! Thanks so much Patrick for this series :-)
Shaun, I found myself asking: how would the biscuit affect tensile strength? My initial thinking was that the additional glue area created by the biscuit doesn't change the glue area along the plane of the failure, but that is pretty shaky. I think your analysis is correct in terms of why the peak force to failure in bending doesn't change. But since the mechanics of wood and glue are so complicated, there is an additional set of head scratching questions. Since the failure here is brittle, boundary conditions are critical: say we move the biscuit as far from the neutral axis into the peak tensile stress, that would effectively mean just slapping it on top and across the joint! That changes the mechanics entirely, since it would be far more difficult for a crack to initiate with the long wood fibers in tension. Where would it fail then? So, while it could be interesting to see the test repeated with a biscuit at different distances from the neutral axis, I think it's maybe more fruitful to think about the broader class of splines, inlays and the like. Each will change fracture mechanics under different loading conditions with their own benefits and drawbacks. It would be fascinating to see different methods compared through tests under a variety of loading conditions: if well done, that could add some real engineering foundation to woodworking practice.
@@jiijijjijiijiij Except that this series is about glue and not about the geometry of the joint. He specifically says that these results only apply to stress as generated by “my rig.” He shows at the end a number of force vectors that are typical but not tested.
@@blacksmith67 you’re right, but it is unfortunate that he went to all this work but then applies the force at a direction that a miter joint like on a picture frame would never have
I do Agrar, tha a biscuit in the neutral axis should not improve the bending strength. In practice, joints with biscuits are stressed often perpendicular to the biscuit plane. That could be a test worth to do. Or biscuits are loaded with a force parallel to their plane, eg. if put in two rows into a thicker board and away from the neutral axis. So I will continue using hundreds of biscuits also in future, and I am confident they will hold things together. Btw: Since I live near Basel where the original Lamello was invented, my English might be only semi-correct
I think the main reason why the biscuits are made in an angle is because you don't know what direction you are going to need the biscuit. Most cuts in woodworking are rips and crosscuts. So having the biscuit in a diagonal covers both options
AGREED 👍
I use biscuits as an easy helper to keep wood aligned, especially on heavy long boards. Paired with pinch dog clamps, biscuits really help when your panel clamps just aren’t long enough or you just don’t want to use them. I’ve never viewed biscuits as a reinforcement but rather as an aid for aligning.
Yes. I use them solely for alignment purposes. Especially when doing long edge grain panels. They aren't perfect but reduce the alignment error greatly and eliminates having to deal with slipping when tightening clamps. The little time it takes to cut the slots and install the biscuits totally makes up for the fuss of aligning panels and the amount of sanding afterwards to smooth out the joints.
conclusion #2 @ 8:10
I thought this was what everyone believed until I started seeing UA-cam videos claiming they don't "add strength." It was always just to keep panels aligned.
I blame Festool.
I'm with you, splines and biscuits are for alignment for me
Exactly, you don't worry about your table holding up an elephant, you worry about board warpage and glue failure over time. Plus mitre joints don't need to carry weight, they need to withstand being pulled apart length ways.
This series is fantastic! Thank you very much for your research and for sharing it with these videos. Keep ‘em coming! 👍🏼
More like it is silly quasi science. First video was pretty dumb and here he tests biscuit joint strength in very silly way, in end grain joint which is weak and unusable in butt joint. Only way to joint end grain strongly and reliably is to increase glue area - in practice this means avoiding end grain joint. In construction lumber never ever butt joint would be accepted, it is so weak .
His conclusions are just stupid
I just watched your glue series and was impressed by the experimental setup that allowed head to head comparisons among the different woods and joint orientations. Avoiding differential calculus in favor of a solid "4X stronger" evaluation was a great choice enabled the design of your samples. Very informative, great presentation, fantastic attention to detail.
I have really been enjoying the conversations you have brought up to the woodworking community. Thank you
Unfortunately you’ve tested these in bending with the biscuits at/near the neutral axis (middle) where stresses would be near zero until after failure of the top/bottom of the specimen. Failure of the top/bottom would largely not be affected by the biscuit at the neutral axis. You’ve indicated the limitation of your force application, but you didn’t limit your conclusions to reflect them. To use your bending setup effectively, maybe consider putting a pair of biscuits so they are located away from the centre of the wood specimen to see if they add strength when actually being stressed before failure. It is nice to see a more deliberate effort to pull apart some of these glue myths, keep it up! 👍🏻
I have to agree with Pat. Your test only covers a case where the biscuit doesn't add significant strength. The other cases that you leave for others to investigate would probably result in a very different conclusion. I would argue that in the real world, a good proportion of the stresses on miter joints probably line up with the "other" cases, and as a result, the average strength of a miter joint would increase with the addition of a biscuit (although I'll admit that a splice is still superiors when it comes to adding strength) .
Your notes about the grain orientation of the wood in the biscuit is interesting. I wonder why the manufacturer does it this way. This brings up an interesting question: What would happen to the strength of a joint if the grain in the biscuit was oriented properly? What if a hard(er) wood was used to make the biscuit? The fact that it tore itself apart during your test means that there is room for improvement in the biscuit...
Thanks for the video, still very interesting!
@@NugTrace I'm tempted to make up a handful of specimens and run them into the lab where I have the capacity to test them in uniaxial tension. Might be interesting... sadly there are so many other things I'm supposed to be doing!! 😁
Agreed. But I also wish he'd modify he's load testing to isolate shear and tension strengths as separate tests. His set up induces a high shear to bending moment ratio that is atypical of actual applications. And a biscuit won't contribute to any additional shear strength since the extra glue area is outside the shear failure planes. I'm not sure how much extra strength a biscuit actually would add, but this test is virtually designed to render the biscuit useless.
@@aaroncecil7509 His short 3-point bending setup indeed, as you noted, make for coincident shear/normal stresses at the point of the joint. If COVID allows, I hope to run a handful of specimens through pure axial tension tests in the new year to see how they will do. I don’t know how they will compare, but at least we will see the biscuit joint being pulled apart. Should make for a fun 4 or 5 min video.
Biscuits and biscuit joiners typically use the centerline. So this test is perfectly valid for the most common technique of how people use biscuits
My understanding was always that biscuits add no real strength, rather are used for part alignment only. Love to see you test dowels and dominos as a comparison.
Agreed. For me the main use of biscuits is for helping align the surface of a panel.
@@mrigeoy That’s what I was taught too.
I'm guessing dominos will be stronger because they're long grain. Not sure about dowels. Probably also stronger than a simple butt joint.
I like dowels. Mainly because it's a cheap system and easier than tenons. I would predict dowels, dominoes, and tenons to have similar abilities because its basically all the same concept
I would use biscuits not for alignment during assembly, but an insurance policy against boards that want to warp in the future, especially trim pieces.
One point to consider: the biscuit was inserted to the neutral axis formed by the bending moment. I think this location plays a bigger role then it's mechanical properties or the gule.
Yeah. I kinda wanna get a tension or shear comparison
or bending moment along the other axis, to simulate a miter opening due to wood movement
@@JohnDotBomb These tests desperately need to isolate shear and tension as different tests. His load testing creates stresses that would not ever be induced in actual joinery applications.
This one for instance induces a high shear to bending ratio (and loaded in the typically nonstressed axis to boot). All the extra glue surface area is rendered useless because it is outside the shear failure plane, and as you said it's extra tensile properties have to impact because it is in the bending neutral axis. Then after the thing has failed he finally puts it in tension when he tries to pull it apart and is shocked it is actually suddenly doing something.
As others have said, alignment is the issue when I pull out the biscuit joiner. Almost always in situations where I have side grain to side grain - not end-to-end grain.
The myth seems to be that people use biscuits for strength since most comments state they have never relied on them for strength. The series does help us all in that it is consistent in method and either reaffirms what we suspect or blows up the folklore of others.
Great series with a lot of lessons.
mitre joints are not designed to resist or withstand the direction of the forces applied in these tests. they're meant to go around corners, so the forces they must resist are those that try to deform them in the (x,y) axes by closing or opening the angle at the joint. this press test applied force only in the z axis, which is not what mitres are designed for. biscuits and splines strengthen the joints in the (x,y) axes, which you proved when you could not pull the pieces apart by hand. i bet if you tested mitres with and without the reinforcement of biscuits you'll find a huge difference.
He is just trolling, he don't care reasonable things like that. It is sad to see how many fall in his tricks
Thanks for articulating my thoughts exactly. I can't think of a situation in woodworking where the stress on a joint like this would be applied from the side.
@@bekanav 100%. It's a shame how many people think this is super helpful information. The information gained from these has no application to what woodworkers are doing. If anyone is making mitres the way he is in these videos they need to rethink what they're doing.
he mentioned that at the end of his video, were you listening? 🤔
@@fabianherrera5427 Yeah, he mentioned it. But he still went on as if his testing was valid. The methodology was flawed, non-representative of real-world conditions. So his conclusions are entirely irrelevant. An accurate test has to simulate a real use case.
Your 3 point bending test illustrates a simple engineering concept for beams. Loaded beams have one side in compression and one side in tension. In between is a thing called the neutral axis, where the stress is zero. In your case, because the biscuit is located in the middle (which happens to be the neutral axis for this cross section), the bending does not even put stress on the biscuit until after the butt glue joint has failed.
I felt like something was off in this test, and I think this hits the nail on the head. The advantages of the biscuit are not well represented in this test. I would have liked to see a tensile test of the miter joint with and without a biscuit to see a comparison. But since the grain direction of these biscuits are 45 degrees, they may not be as helpful as I thought.
This is a good point, but the neutral axis is an infinitesimally small line. Biscuits are not that. It is possible that the biscuits break even with a small fraction of the stress that's applied on the outermost surfaces before those break.
@@RishabhGKoenigseggRegera The test wasn't about showing the advantages of a biscuit. The series is about the glue.
I was going to mention this. Most of the strength of any structure for most types of forces comes from the outer surfaces.
If you're using it close to the limits, you could epoxy much strong dowels deeper into a joint, like steel rods.
You’ve put a lot of time and effort into this project, and I appreciate it. Thank you and God Bless all that you do.
Thank you for the knowledge, care and effort you poured into this gift. Your videos are required viewing from now on.
I appreciate that you point out lignin's comparative weakness as an adhesive. PVA is a modern marvel.
I especially admire, and appreciate how you focused the content so that each video could be limited to ten minutes, the optimal viewing time for retention.
Of course, the points made raise further questions, which underscores their value. I can't wait for the next one.
Thank you for all your time and hard work, Mr Sullivan.
That sales ad was hilarious. Keep up the great work, good sir!
Thanks for sharing your very thorough and logical research. I'm new to woodworking and as a retired layabout have lots of time to watch these myths being put forward in various ways. Your three simple videos are incredibly informative and most reassuring. I'm not an engineer, just enjoying the beauty of wood and what you can do with it, these videos are brilliant.
Great videos Patrick. I know some others have suggested it, but it is "common knowledge" that domino joinery is just as strong as a traditional mortise and tenon. I've always doubted this, but I would love to see a test done. Keep the videos coming! Woodworkers have been doing it right for millennia, but technology has changed and anecdotal truisms can be tested!
I watched a video about what makes the material strong when forces are applied like this. One surface is under compression, the other is under tension. Most of the material in the middle does not contribute too much. So the only way to increase strength is to put something on the surface. even thin veneer will do better than biscuit. However ff you are making a frame a spline will add to the strength, because as you said there are other directions that forces can be applied from.
Mr. Sullivan: I very well done video....the filming, voice over and graphic work is top notch. Thank you. I would only say one thing though about the content and that is, we don't use biscuits for strength, we use them for alignment. If I'm trimming a window and using 1/4 stock or mdf flat stock or whatever I may use a biscuit so that the faces align.....not because of strength. The glue is strong enough on it's own. I trimmed a house once and after finishing the lady decided she wanted shiplap everywhere so a guy came along and took down the trim around most of the windows and doors. They were coming off like picture frames ...in one piece and they were only glued (at the miters). Glue is plenty strong on its own.
I love the way you explain things so that even the absolute beginner can still fully understand. Well done
Hello Patrick Sullivan.
As a mechanical engineer I can tell you why a biscuit wouldn't add strength in your test, even if the biscuit was made of something much stronger.
Whenever a given object is experiencing bending, the actual stress in the material is NOT uniform across the thickness of the object.
It is the highest the further away from the center (oversimplified) you are.
And as such, a biscuit put in the center of the boards won't ever get to help reinforce the joint when it comes to bending.
Perhaps a test with a board thick enough to have two biscuits in the height plane could prove that biscuit either add or subtract strength.
The biscuits would need to be close to either surface instead of bundled together in the center.
So perhaps a 20mm thick piece with either biscuit being only 4mm below the surface would prove to be stronger or weaker than without.
I think this test shows the limitations of your set up. It's worth doing these tests again but instead testing the pure tension strength of the glue joints by seeing how much force it takes to pull the joint apart
This series has me glued :)
Cant wait to see how dowels hold up!
@Vikas SM yes, I’m curious about this as well.
@@ryanlemons7831 I mean, there's a reason dowel joints have been used as far back as we have any kind of record. There isn't a surprise waiting there.
Haha, you had already addressed my thoughts concerning the previous video before I even thought of it. Great video! Very useful and informative, as the previous ones. Thanks!
You nailed it when you state the area vs the direction of the force applied equal the strength of the joint. With a biscuit, the area of glue is actually DECREASED when you are talking about the direction of the force applied. This means the amount of strength gained is about equal to the amount lost to the lack of glued surface in relationship to the direction of the force applied.
Patrick, Love it. I would also love to see results in pure tension and shear too, if you could do it. The bending moment butt joint to failure is actually a combination to the two. Envision the top of the joint is in tension and the bottom is in compression over the fulcrum. Also, try smearing some glue as deeply as you can by rubbing it into the end grain and then adding a bit more glue. I've done that on miters in red oak and it seems to be even better. Did kitchen cabinet doors like that (no splines or biscuits) 30+ years ago and they still are together. Thx for the testing!
It was my understanding that biscuits were for alignment and a bit of strength for butt joints (ie. cabinet rail styles, frames etc) or slab glue ups. Very interesting info, thanks!
I made a solid oak table, used biscuits to help align the oak planks for the top, it has never cracked or separated along the joints. Later I needed a quick T.V. stand, used biscuits as shelf supports, glued and clamped. let dry. That stand held an old heavy 27 inch tube T.V. for a few years before T.V. was replaced and is still holding together.
Another great video, can't wait to see what your next test involves.
This series is *fantastic*! I hope you continue it to include other types of loose tenon joinery such as traditional Dowels, Dominoes, and Beadlock!
The purpose of the biscuit is for alignment during glue-up and to hold the project together even if the glue joint fails. Also, biscuits should be used with a water based glue to work properly. Biscuits are compressed and the water is absorbed by the biscuit making them expand to provide a tight fit and proper alignment. I've glued up biscuit joined pieced and then needed to pull them apart several seconds later because I goofed up and they would not come apart because the biscuits swelled up and held everything in place. I switched to Titebond III and noticed that my parts still had a lot of movement to them even several minutes after being glued. The reason being that itebond III is a polymer based glue and if it does contain any water, it's not enough to swell biscuits.
Helpful! I've been resisting buying a biscuit joiner, although I've used them before. I know they have their place, but I think there are better alternatives. Thanks for this video!
Biscuits are great for aligning pieces to eliminate "lipping" and making it easier to clamp mitered frames together as they minimize slippage. So they do have a use, you just have to understand that when deciding whether or not to use them. Same with floating tenons popularized with the festool domino
As someone else pointed out, the biscuit was inserted at the neutral axis. Before the test was conducted I expected the result would show no difference from previous tests. In a bending load the max stress is at points furthest from the neutral axis, which means that peak load depends on glue strength, and the biscuit's contribution is negligible. A complete set of loading conditions need to be tested before meaningful strength conclusions can be made. However, even with that information, I doubt whether any wood joinery design would be altered.
As always biscuits are tested in a manner that is completely unrelated to their design purpose. I think the general public really misunderstands this tool.
Another great vid.
This vid series really breaks the myths around different glue joints.
Those other force directions that you leave for others are the primary way that miter joints are used!
The way he's testing them is the worst-case scenario though, i.e. the direction that miter joints are weakest. So I think it's okay.
First test was a bending test along the minor (short) axis. The peak stresses would be at the edges with only the glue bond. These stresses would be the same with or without a biscuit. A tension test or bending along the major (long) axis may be more telling for just the biscuit strength itself, but still won't help the table strength with end grain connections.
Conclusions about fit-up are good case. I know people that just use dowels over the biscuits or dominos. Be interested to see if dowels are any different than biscuits.
Also the cabinet shop I worked in as a young man we would use 23 gauge headless pin nails four on each side to bind the biscuit and joined wood four pins per side.
I'm really enjoying this series - I hope you'll be looking at mdf, plywood, etc. at some point.
Thanks! Very good analysis within your test perimeters.
Pat H has identified the exact weakness of your testing method. Think of a steel I beam. The top and bottom flanges are wide in order to put most of the material (strength) where most of the stress is when a bending load is applied. A very small thickness of metal is in the center of the beam (web) the stress from bending is lowest along this line (neutral axis). There is no location in the joint where the biscuit contributes less strength to a joint in bending.
This kind of testing is important so please keep up the good worl,
Keep up the good work! I'm really enjoying this series!
Going back to old Norm Abram material (who obviously deserves a lot of props for making this possible for so many of us to discover, so bless him for that) from the late 1980s, there was definitely a time when we thought biscuit joinery was almost magical. It’s amazing how eager we are to think that of new things.
As usual, these tests in the end are most of all a reminder to think through what our assumptions are and ask if they actually make sense in the context of what we’re doing.
Thank you for this outstanding series.
Looking out for more!
I suspect it's not so much a matter of "glue area doesn't matter" as "glue area where the stress concentrates does matter". It's the difference between static and dynamic analysis (ouch, using the Wayback machine to 1970!). Any force applied becomes tension in one area and compression in another, and the exact location of the critical stress can be deceiving. The location also moves as the stressed structure deflects under load. The 45 orientation of the biscuit grain is probably a clue (manufacturers didn't arrive at that orientation by accident). Biscuits might contribute more to joint resistance when the joint is loaded on a different axis than the simple end grain reinforcing axis. However, at least it's good to know biscuits don't reduce the joint strength. Excellent videos Mr. Sullivan. And to your profession: "thanks for all the fish". (I'm a forensic engineer!)
Another fantastic informative production
Been waiting for this one! My guess is that the glue joints of the biscuit to the slot don't fail - the glue area that counts is the butt to butt joint is actually lower, and the biscuit itself cracking is proof that it's adding about the same resistance to the glue lost, not any additional. i think that's my takeaway from your testing too... and i think the point about facilitating assembly and getting good alignment between parts is right on the money. either way another great, data driven video that had my riveted
I bought into the biscuit joinery fad back in the 80's and spent a hefty amount on a Lamello, it took about 6 months for me to realize that it didn't improve my work or provide any benefit to what I was doing so I sold it and never looked back.
Hang on. You tested bending strength about the centre of the joint. Had you tested pull out tensile force then biscuits would have reinforced the joint. Also biscuit material/ strength varies. I’ve used some quality biscuits which do most certainly add strength. A few too many simplifications I think in this video.
I have always thought of biscuits as a way to “locate” joints. Since I have learned the technique of dry-fitting joints with blind pins, I no longer use biscuits to avoid slippage.
Thank You for your excellent video
I seldom use my biscuit joiner except for difficult joints never for strength.
Small tip for biscuits keep them in a sealed bag or container add descant pouches for added protection from moisture.
Biscuit/plate joinery in many shops is mainly to speed production and glue up, not necessarily for strength. I've always thought of biscuits as strong though, generally.
Excellent video. Appreciate you sharing your tests and conclusions.
Another excellent review of a wood glue joint consideration and examination. I which you would do the same test with 2 or 3 dowels in the joint vs biscuits since you can do the testing and I cant. Hope you do to see the results on that set up. I have enjoyed your video and these studies. I have never understood biscuits and the reasoning behind them. 5 thumbs up from me Patrick! Columbus Michigan
Patrick, thanks for this clear and understandable video series. I hope your viewers' take-away is that woodworking joints should be designed to rely on the interlocking structure of the joint for their strength against the forces at play during the object's intended use. Glue is intended to prevent the joint from slipping apart in the reverse direction of assembly, which should be in the direction LEAST likely to have excess force applied during use.
I use biscuits to keep things aligned in clue joints that I suspect will break.
For example on a chessboard along the endgrain. I know from building many chessboards in the past that it will have between ⅛ to¼ inch in movement so there's going to be an endgrain joint somewhere that will fail. The biscuit is there to keep that joint aligned and intact.
Great series but as you pointed out in the end, you didn't test the most common problem forces.
Indeed, those test are interesting but poorely designed and naively interpreted
Could be the name of this series. "Force tests on wood in ways never used"
Gosh, I’m learning so so much from your videos. Thank you thank you thank you!!!
Really love what you’re doing, fantastic data. I was always taught to keep a open mind and never stop learning. Thank You for this series looking forward to seeing more information!!
I make my own splines for the joints and use a biscuit joiner to cut the slot. The pieces are easier to line up.
Very nice, factual presentation. Keep them coming, please.
Hey, do you think you could make a video like this testing dowels?
Patrick, Your series on joint strength has rattled my long held beliefs. Very interesting and makes me wonder about the double tenon joint strength. Festool is a common system that I’ve believed to be superior to the biscuit type joints. Can you please do a series on their strength in one and two row configurations? Thank you for your exceptional videos and please keep exploring myths/truths about best construction techniques and possibly glue types and properties. Cheers, Will
In think leverage is everything and limiting it with grain orientation is the most important factor. No matter how strong the joint you can break it easier if leverage is available to the joint because of the finished project allows it.
My favorite application for biscuits is attaching a face frame to a plywood carcass; biscuits have pretty much entirely replaced brad nails for me in this application for several reasons, including biscuits are stronger than brads while the glue is wet, and there are no nail holes on the finished exposed surface of the project.
Here's one thing I wonder: would adding a biscuit (or any other spline) increase the strength of a miter joint not by adding its own strength, but by helping align the joint and thus forming a better glue bond? Probably not in your test rig where a "miter" joint is mechanically a butt joint but in a real "two boards with 45 degree ends meeting at a 90 degree angle" situation. Those joints are difficult to assemble and clamp, so adding a biscuit or spline to constrain the two boards together may allow for a superior miter joint.
It's one of those things where the "I made some joints and then used a bottle jack and a bathroom scale to break them" tests kinda fail; I think this methodology eliminates variables that actually matter in practice.
Brilliant demonstration and video. Thank you.
Really enjoy your videos, would love to see you test dowels
If you want to treat a biscuit like a piece of rebar then you would add the biscuit as close to the outside edge as possible. Adding strength as close to the center as possible does little to prevent bending forces.
This is actually an important takeaway 7:16. This test, like many others have said, did not test the type of failure that when used in a miter joint. The test and its conclusions are valid though in the case of panel gluing, due to the biscuit being placed in the neutral axis of bending.
I would really like to see a version of this test with dowels.
Excellent demonstration! I wonder if dowels would work better. I guess that would depend on how deep the dowels penetrate into each piece.
Ha ha, I knew or lets say felt that biscuit joints doesn't do any much help to strength :) so I never bought one and I always thought they look stupid... But it seems you have more opposite comments for this video compared to the first two videos and I guess that is because many new wood workers have been convinced to buy a biscuit machine and using them since a time and don't want to accept that that machine is not much use except for the aligning purpose.
But for my view, If you cut the 4 pieces identical to each other, there will be no problem with alignment and you can also always hand plane them a little and there you go, you have a perfect miter...
Thank you very much for the video 👏👏👏👏
In the shop where I work, biscuits are used primarily for alignment. We build drawer boxes out of melamine by placing a biscuit or two in the corners of the drawer box. These joints also get glue on the spline joint as well as melamine glue on the butt joint itself. Finally, the joints are stapled together. We never have any failures with these joints. I have taken some of these drawer boxes apart and they always break apart with the particle board still glued to the biscuit. In this case the biscuit is stronger than the particle board around it.
I was given a large divided cabinet, melamine covered plywood. all the dividers and shelves were biscuited, I had to cut along the shelf gap to get apart. the glue hadn't stuck to the melamine but the biscuits sure put up a fight. When I did get some shelves to come out, the plywood/ glue failed around the biscuit or some just broke. Was crazy how hard it was to take apart.
I use biscuit to align, and also for use with melaminé, which have a plastic coating on it.
In the direction that you were testing the samples, the pressure on the joint would tend to split the end-grain joint on the OUTER surface first, and then very rapidly the split would tend to progressively work it's way into the biscuit (this is milliseconds fast). Then the moments acting on the straight grain of the biscuit would tend to split the fibres apart, as they rotate, to try and reach the direction of the pulling force acting on them, thus tearing the biscuit and the joint apart. So it seems likely to me that the best thing to do, is not to put ONE biscuit on the centreline of the thickness, but to put TWO biscuits in, just under the surface of the back and front faces. That way the biscuit will act like plywood just below a thin veneer. I wonder if you tried that, the results would be different???
Excellent work once again Patrick!! I had watched part 1 and 2 and wanted to see the spline effects. With my Mechanical Engineering background, I have to totally agree with the comments about how the force is being applied, i.e., along the neutral axis, and that is not the practical application. As most everyone has stated, the biscuits, dowels, and dominos are all used for alignment while gluing.
It would be interesting to see what would happen if there were (2) biscuits evenly separated away from the neutral axis, but I don't think your test sample it large enough to allow for that? However, could you turn all cases of your test pieces 90 Degrees to simulate more of the breaking force that would pry open the joints over a longer distance?
I truly enjoy and appreciate your work. It is all accurate in the approach you have taken, but like prototypes, we learn new setups as we go and come up with ultimately a better product. Thank you very much for providing data versus just opinions, which everyone else can easily do. LOL
If there is any movement in the glue at all, the leverage on that biscuit is massive.
I made a mortising jig for my router. It uses a bushing that rides in a slot at the top. I based it on Phil Morley's jig, but I made it so I can cut pieces of all angles (though the top has to be flat up against the bottom of the base), so my jig uses t-tracks and blocks with clamps on them, so I can do very narrow pieces (down to 1/4" mortises on 1/2" pieces of wood) or and I can do very wide mortises, and/or cut into pieces that are very wide, so the mortise can be set back far from the edge. The top moves about 6", so I have a wide range of choices... I then custom make loose tenons to fit into those mortises. This makes me want to test the breaking strength. I like the loose mortise and tenons because the mortises are fast and easy to cut, especially now that I've designed and built my jig, and the tenons are also very easy, and can be cut with repeatable accuracy. I feel they are better than dowels because they should resist any racking/twisting more than dowels, and part of me wonders if they may be structurally better than hand (or machine) cut mortise and tenons that are not the loose kind. They certainly provide much more structural stability than biscuits. I built a gate this past summer, when I replaced my fence. The gate is 9'6" wide, and in the middle is a W, with those four straight lines of the W each being a piece of lumber, with a loose tenon attaching them together, and I was able to lift all four 2x4's with the tenons in place, and the W held its shape, and that was before I had having put any glue into them, and before I had put the W into the frame of the gate. In any case, I appreciate your expertise - you clearly have more experience with this than I, and it is interesting to think outside the box, and you did, literally, break the mold, so thank you for that!
Great series! Any plans to do a test of dowels?
Watching your series, I am fast becoming a fan of my own choice to use lap joinery in my artistic panel's cradles...
In my opinion, when seeking to add strength first understand where (the direction) your joint is expected to experience force, then look to how you can strengthen.
I use strong dowels at 1/3rd the thickness of the thinnest board being joined.
Very informative. Thank you for going to the trouble to present it.
Please also add splines and dominos to your test sequence!
This is very informative!!!
very good work. i feel i need to wait the end of these series of videos de make conclusions, once all de preparation for each video is so laborious and there for takes time. also i see you are building a train of thought and are wise enough to make the choice for each subject. i am very pleasy for your work. thank you a lot.
Excellent educational videos. Thank you!
However, the LINEAR strength of biscuits adds to the linear tensile strength.The glue joint has 2 axis lines, and the pull strength of this joint is actually improved. Then again biscuits also align the joints so glueup is easier and faster.
I wonder if maybe the reason the biscuits themselves were cut on a bias was solely to decrease breakage and waste in the manufacturing process.
If they cut them with the widest part of the biscuit as end grain (the theoretically strongest way for them to cut them as far as the joint is concerned) I'd bet that alot of them would break in half just tumbling around conveyors etc.
If the cut them the other way they might last longer in the factory but would be actually useless for the end user.
The 45 deg cut is a compromise.
Oh man, oh man, another episode! Can’t wait for the juicy debate to follow. These videos have shaken up the UA-cam woodworking community in unexpected ways like publishing the special theory of relativity or early behavioral economics!
Using biscuits to add more gluing surface to a joint is like cutting your boards in half length wise to add more gluing surface. Technically it will add surface, but it doesn't add any strength to the joint. I've always viewed Biscuits as a device to help align boards together. If you need additional strength, a half lap or tongue and groove type of joint will probably be better.
Another excellent presentation! Thanks.
Very helpful and informative. thanks
Thanks to a biscuit-joiner esque tool from Grizzly, I've switched entirely to dowels with most of the efficiency of biscuits but greater strength (as far as I can tell so far
I would like to see a push test since ussaly I use splins to help with wood movement not down ward pressure but out war pressure
I'd imagine that a biscuit would add significant tensile strength to a but or miter joint, and ultimately most forces applied to a miter joint would stress it along it's tensile axis rather than its sheer axis. Maybe there's a way to devise a test contraption that will pull apart a miter joint perpendicular to the glue line?
The results surprised me. I expected a small increase in the joint strength with a biscuit, but I still expected the biscuit joint to be weak. Two of the things going on here are:
1. The biscuits should be installed on the outer edges if additional. That would be ugly, so I'm not suggesting that this be done.
2. The direction of the fibers of the piece used to strengthen the joint need to be across the cut. The problem with a butt joint is that there are no wood fibers that cross the joint. Biscuits would help more if the grain direction of the biscuit were changed to be parallel with the grain direction of the wood.
The bottom line is that almost any normal wood working technique would strengthen a butt joint more than a biscuit. Biscuits are bad because the are thin so they concentrate the strength they add almost entirely in the center of the joint. Biscuits are bad because they grain direction is wrong for a butt joint. And biscuits are bad because they add less glue area to the joint than most other wood working techniques would.