Thanks for shout out. On the test I did, The long grain to long grain is actually a cross grain connection. So they should be fairly similar in that. but with the rig I was using, it puts more pressure on the glue joint itself. I would to try it with your computerized laod cell. That is so sweet. Nice work. I always love watching your tests. Lots of great analytical thought-through on that.
Hi Matthias, I work as an engineer for one of the top 3 adhesive manufacturers. What you measured is the torsional shear strength of the wood glue. One interesting thing that you pointed out is that the different failure mode in each sample. You can see the "adhesive" failure when the bond between wood and glue failed leaving a clean surface. Most appeared "cohesive" meaning that the glue failed in the bulk structure leaving adhesive on both sides. The wood breakage was a substrate failure and these results are typically discarded. A factor to account for when comparing to Wood by Wright's data is that your setup includes leverage from the force applied at the end of the wood block. Also important is each glue's reliability ratings like humidity and temperature that you mentioned. More rigid glues like superglue also might not be able to and the CTE mismatch during temperature swings.
@@DakotaBurgener Entirely depends on the substrate (wood or other material) and operating conditions (exposure to temperature, humidity, force, time etc.)
I was following you fine until the final sentence: _“More_ _rigid_ _glues_ _like_ _superglue_ _also_ _might_ _not_ _be_ _able_ _to_ _and_ _the_ _CTE_ _mismatch_ _during_ _temperature_ _swings.”_ Might be able to what? And what is a "CTE mismatch"?
@@BenInSeattle coëfficient of thermal expansion. Most materials expand if temps go up, and contract if temps get lower. The amount of geometric change per unit of temp change is the CTE. So each material has its typical CTE. For joint strength stability it is benificial if CTEs of glue and wood are as similar as possible. If not, then temp change causes internal stresses in the joint, as there is only a small temp range where the geometries closely match. Cyclic stresses can cause microfracturing and debonding within the joint. Leads to strength loss and ultimately to failure of the joint.
@@seigeengine I agree with that. A long time ago I pointed out PF was doing biased comparisons and the indirect reply just confirmed it - never watched another video since.
I'll pop in early and say that you used the wrong construction adhesive. It has to be PL Premium - a polyurethane construction adhesive, and not the normal subfloor adhesive you tested.
@@F0XD1E I never said it was as good as wood glue, but it's good enough and certainly good for woodworking. My 20+ years of using it are proof of that, at least for me :)
Fun fact about wood gluing my late Great Uncle Ralph taught me back when I was a child: (Using plain Elmer's wood glue circa 1986) when applying wood glue, brush it on all surfaces being glued, and let it cure for about a minute, then apply a second coat quickly and then put the pieces together, the extra tack and time for the glue on the wood helps evenly penetrate the surface of all the pores in the woodgrain, also helps the glue to remain wet longer when you reapply, thus giving it a longer more even cure rate in the end, and will not "flash cure" (when the wood wicks away the moisture of the glue too fast) thus weakening the glue's bond into the wood. Moistening with water will only dilute the bonds made on the surface of the wood, so that will not work. I still have the example of this very lesson, a desk I made for college, which underwent no less than 6 moves cross country, and none of the joints have failed to date.
It also helps a lot to apply the glue, brush it onto each piece, let it sit for a minute, apply more glue, then rather than just put the pieces together, SLIDE the pieces against each other a bit until they naturally seize together, then slide them into the correct position. I don't know why, but when this sliding technique is done correctly, it makes the connection much stronger.
@@matthiaswandel Universal truth of superglue... it always ends up on your hands. One thing to factor in as well is wood expansion/contraction breaking brittle glues (like superglue) over time. Construction adhesive can be much stronger e.g. PL Premium x8 and MAX with rated tensile strength 2-4x that of regular ones.
@@elminz A compliant glue like polyurethane is better if the temperature is going to vary a lot, especially if the two parts that are glued together are different materials. Old lab trick: if you want to separate two pieces of different materials that have been glued together, toss the assembly into liquid nitrogen.
Matthias, the amount of content you put into one 13 minute video other channels don't manage to put into 50 minutes. Thank you for always cutting straight to the chase!
I use construction adhesive regularly for, well, construction. What I’ve learned is there are many different kinds for specific purposes. DAP4000 for example has very high PSI ratings for cross-joint strength. So when your subfloor expands and contracts seasonally, you don’t shear the glue off. I imagine that you used one of the glues that’s designed to withstand expansion, and so is weaker (but less brittle). The more brittle stuff might make better furniture joints
Yep, I'm in the states and unfamiliar with LePage, so looked up the one he's using and it is for subfloor. LePage and Loctite are basically the same it looks like. Same parent company and similar brands between them. The PL 3x and 8x did really well in Project Farms testing
Thank you for this! I never even considered gorilla glue before. Three factors for choosing titebind 3 Vs regular wood glue: 1. Outdoor/waterproofing needs 2. +5/10 minutes working time 3. Remains non tacky for longer, which makes bringing several tight joints together easier (like a row of dominos or finger joints).
As always, fun to watch your tests. Loved it when the samples popped off and flew through the air! I've been using the Gorilla wood glue for the last few years and have been quite happy with it all around. Nice to see it showing well in the standings!
Here in California, at least, "construction adhesive" used to be a thick tacky paste that smelled strongly of hydrocarbon solvents. It worked well for large gaps and disparate materials, but the law now limits VOCs, and all construction adhesive is basically latex paint, which is slow drying, and weak. I agree that cheap wood glue is still the best for tight fitting wood joints. Thanks for doing the tests!
With the -20 to -30 degree weather we've been getting here in Toronto....my LePage wood glue literally froze into a block. The woodworking never stops though. Started leaving them in the house but sometimes I forget the bottle in the garage.
Good test. The Titebond III is waterproof though. Don't know about the gorilla wood glue. A few percent isn't a major concern to choose one over the others for all applications. The top performers are all fairly strong and are likely to break the wood before the glue. Stuff like working time, cleanup, water resistance, finish appearance, and sandability all come into play for real projects.
Matthias, thanks for doing this test in such a controlled way. Multiple samples, and then glad you went to hard wood! Very surprised that CA glue did so well.
As a retired custom furniture/cabinet maker, I've used a lot of glue, and I'm still doing some woodworking. I got a kick out of how your "test" came out compared to my favorites. For ME, construction adhesive really hasn't been all that strong, and my two by far favorite woodworking glues are, Gorilla wood glue and Tite bond 3! Looks like we are on the same wav length! lol SR
@@marcelo403polo2 Titebond has been my go to for home audio projects for years, this makes me want to try out the GG to see how that works out. I’ve also used construction adhesive for a few larger 30-40 cubic feet enclosures and I’ve found that it’s got a long set time so positioning awkward pieces was always easy. It fills gaps great and it sticks to most any surfaces + angles.
@@Noughtta Being that it's formulated for structural use, I wonder if construction adhesive has more give to accommodate building flex without cracking. If so, that give might have an impact on the acoustic properties of the enclosures.
Titebond 2 is my go-to, especially if you need something that sets up quick and for most interior projects.. Titebond 3 for exterior stuff or when you need a longer working time!!! All the other choices are garbage except the cheap "wood glues" which are fine for most DIY/weekend warrior stuff. Construction adhesive is just that. Construction adhesive. Yes it will hold but no where near a real wood glue. My fav construction adhesive is probably Sika adhesive, the gooey PL stuff is pretty good too but I'll always grab a tube of Sika before anything else.
As much as I crave another one of your multi-part builds (like a marble contraption) the glue test was riveting and very informative. Without seeing this, I probably would never have even given Gorilla wood glue a chance which sounds like it deserved, so thank you Matthias. And don't stop leaving them uncensored, real reactions are well, real :)
Great test, thank you. A very useful test would be to glue up your samples and let them sit for a year, then test them again to see how many of them have maintained their strength over time.
That's a great comparison! Thank you. Further ideas for testing; perhaps try wiping down the gluing surfaces of the softwood with acetone or something to try to get rid of potential oiliness? And/or maybe scuffing the gluing surface a bit with some 80-grit sandpaper (even though it was already rough). Or maybe the softwood is just so relatively flexible that the weak points in the wood/joint are allowed to move enough for a break in the glue joint to propagate more easily.
I would like to see a similar test with the pulling apart strength. This 90 degree test is good but introduced a twisting effect to break the wood or the glue. So, to see how it pops apart too to bottom not sheer strength.
Thanks for doing all those tests Matthias. Very interesting to see all those results. That must have taken a long time. The construction adhesive that you used is different than the one John uses which is a polyurethane. I use the same type that he uses. I was building a workbench a few years ago using that polyurethane ca and realized I had put a piece of wood in the wrong spot. By the time I had a chance to get back to correct my error it was about 20hrs. I was banging that piece of wood with my framing hammer multiple times and could not get it off.
yes. the pl premium is the polyurethane based glue which cures with moisture. i'm not sure how that pl400 cures. I would like to see how the pl premium holds up, i'd put a dollar on it winning or tieing with the epoxy.
Bravo for the constant strain rate tester, a DIY Instron. As with wood glues, there are numerous construction adhesives. Not all of either category are created equal. The Liquid Nails construction adhesives tested by Project Farm in July 2019 was at the bottom of the pack.
Love these testing videos. Been watching your vids for years, and all the testing videos you've ever done are always my favourite. I just love cool graphs!
Then there’s longevity. Not only is construction adhesive weak at the outset, it decides to let go years later. Our contractor glued and screwed floor decking to the joists. “You’ll never have a squeak,” he pronounced. A few years later, the adhesive had all let loose and I had to install many, many more screws to quiet the floor. I wouldn’t use construction adhesive to build a henhouse.
I feel like the surface finish has a lot to do with how good the glue holds. Something like oak with a ton of texture and pores combined with a good runny wet wood glue (water based to open up the pores too) would hold 2x as well as on a freshly jointed maple face which is ultra hard, and smooth and non porous.
My general use shop glue has always been Titebond 2 and it does go bad after 4 years or so Thanks to you, Matthias, I will keep some Gorilla glue on the shelf!
The other problem with Gorilla Glue is its shelf life. Everytime I have ever tried to use Gorilla Glue, it had hardened in the bottle before I could ever use it. I eventually stopped keeping some around.
That's what he was referring to with the small bottle of Gorilla Glue (the "original"). The Gorilla Wood Glue he recommended at the end is a totally different product.
TB3 is still my favourite as a general use wood glue. I've never been a fan of WB or LePage in the past. Gorilla glue always worked as a "repair it now, not repair it right" sort of solution. Good to have around in emergencies but certainly not a woodworker's staple, in my humble opinion. Just like construction adhesive it's great at dissimilar material bonding for funky applications. The thing that stands out the most is that the normal cure two-part epoxy is weaker in your tests than the 5 min. I guess your data might warrant looking into LePage again. Thanks for sharing, good sir!
I'm thankful that the winner in this case is actually easy to find almost everywhere. Given the results with CA, now I'm curious if spray on CA will yield similar results. Having a very strong spray on glue for wood could be really handy to have around a workshop.
Great follow-up to the last video. I was going to suggest doing a time-based test as well, to see how the glues deteriorate in strength over time (e.g. the super glue), but then you did the test with the old gorilla wood glue which was very interesting.
Fantastic work, Matthias! Thanks a bunch for all the tests! 😃 Hot glue is pretty impressive indeed! And no one seems to trust it. Go figure. 😬 Anyway, stay safe there with your family! 🖖😊
That's how most PVA glues will behave as long as they're stored properly. Reputable manufacturers will always put a date of manufacture on the bottles though, and have a recommended shelf life. TB wood glues are 2 years minimum, so they're what I stick with.
@@NWGR if you store them slightly under the MFT / Chalk point (according to the data sheet) it will extend the life of the product usually dramatically
At work we did some material testing and we figured that the adhesives are still curing after 1 week. We saw an increass on performance by ~20% between 24 hours and 14 Days.
Your previous video where you experienced the bridle joint glue failures had me wondering...was it really a glue *failure*, or was it glue *starvation*? With the testing here, you just did a single face joint, which allowed getting a nice even layer of glue between the pieces before clamping. But with a bridle joint, or even more so with a box joint, you have to slide the pieces together during assembly, so it's not really an apples-to-apples comparison. I have often wondered if having a tight-fitting joint before adding glue just ends up scraping all the glue out of the joint during assembly. Your bridle joint tests where it appeared there wasn't any glue applied after they failed seemed to reinforce that theory. What might be an interesting set of tests would be to try bridle joints with varying clearances and use the same glue for each to see if the problem is that you are pushing all the glue out rather than the strength of the glue itself. And if so, what is the optimum clearance before the glue layer becomes too brittle and weak due to thickness?
Interesting point. Matthias once said of the foaming Gorilla Glue, _“I_ _guess_ _it's_ _a_ _good_ _glue_ _to_ _expand_ _into_ _inaccurate_ _joints,_ _but_ _I_ _prefer_ _to_ _just_ _make_ _my_ _joints_ _so_ _that_ _they_ _are_ _accurate.”_ If Matthias is accurate to the point of nearly zero clearance, it wouldn't be surprising if he scraped the glue out of bridle joints.
Great video, never would have thought CA would have done that well. Real world feedback. Gorilla foaming glue doesn't hold up in the humid Texas Gulf Coast climate. I used it to repair part of a fence picket (the top piece broke off due to a knot and a utility line worker). It lasted 3-4 years before the glue failed and the wood again became loose. I cleaned it the best I could and used Tight Bond III and so far (4+ years) its holding.
I think you cleared up a mystery for me. When gorilla glue first came out, I picked some up because I had been impressed with their duct tape. I had this same experience, it bubbled up like a foam and separated the joint. A year later I mentioned this offhand to another woodworker I know, and he's where he had been using nothing but it for a year. He had no idea what I was talking about when I said it was foamed up. I was completely unaware there was a second type of gorilla glue that was not that horrible stuff. I normally quite agree with this guy so it was one of those things that stuck with me, off putting.
Gorilla is a BRAND of glue, not a type. They make Fast and Slow set Epoxys, Metal-reinforced Epoxy (aka JB-Weld), Contact Adhesives, Super Glues, a full range of glues. The bubbling stuff is Polyurethane glue, and it's also available from many other suppliers (eg: Sika). If you read the instructions, you will see that you MUST clamp that style of Polyurethane glue. You know, RTFM or the label on the bottle.
The adhesion of the glue to the softer wood is better than that of the hardwood, however, that bond works against the strength of the test not for the strength of the glue, but FOR its strength over the wood, the wood is sure to fail faster, most of the failures happened due to the rigidity of the glue, and the lack of strength in the lignin bond between the layers of wood.
Looking at your videos of the "glue failures" situations (as opposed to when the wood itself is cracking and splintering), it looks like the specific failure most of the time is in the adhesion of the glue to the wood surfaces. I would conjecture that CA glue is superior because its lower viscosity and thinner nature allows some amount of penetration into the depth of the wood, rather than just staying on the surface, thus forming more of a 3-dimensional bonding zone between the two pieces of wood. Think penetration in metal welds as an analogy. If this is indeed the case then it would be very interesting to see if macroscopically roughing up the mating surfaces, to make the bond zone thicker, could increase bond strength with some of the weaker adhesives.
I seem to recall seeing one YT tester saying that a smooth surface created a better bond than a rough one (which was memorable because I always thought the opposite). Anyway, that might explain your softwood vs. maple; you even said the softwood was freshly sawn and quite rough.
Thank you for this presentation of wood glued, was thinking of getting weld-bond glue and it seems that it looks really good on the graphs without hot glue or epoxy
Thanks so much for doing this Matthias. Gorilla wood glue has been my favorite general purpose wood glue as of late and I’m glad to see a good showing here. Titebond 2 would be my #2.
THANK YOU FOR THIS!!! I have the big chug (3,78l) of titebond III since almost 6 years ago and it still works great !:D I stored it good though, in the cellar.
I mainly use the lepage glue. I have dealt with a panel glue-up that was the standard foaming gorilla glue and one of the lepage. The gorilla glue failed on the seam and it was the glue that failed. The lepage worked so much better and never failed under the same conditions.
My wife got me some Gorilla wood glue from the Ambassador Bridge border store for real cheap and I am very impressed with its performance. I'm glad to hear it still works well when old. I agree that there is no "one" glue. It all depends on the purpose at hand. Gorilla hair-gel (polyurethane glue) is great when gluing wood to a non-wood surface, for instance. Or of course when you're having a really bad hair day. Titebond III is the choice for cutting boards where water-proofness and food safety are concerns. I use hot hide glue as well as fish glue for stringed instruments. It would be interesting to see a test like this performed with those. Great video as always, Matthias. Keep up the good work.
I would really love to see similar tests with threads tapped into a few different species to see how strong wood threads are vs using threaded inserts and T-bolts.
I'd be curious of the differences between Titebond 1, 2, & 3. My understanding is that Titebond 3 trades off some strength for flexibility, so when it's used outside it has some give as the wood expands. Whereas Titebond 1 & 2 are stronger, but more brittle.
@BreatheScotland Titebond 3 is waterproof. Titebond 2 is water resistant and titlebond 1 isn't. Gorilla wood glue(big bottle here) is basically the same thing as Titebond 2.
@@bobbyt9999 yes for indirect food contact meaning the joints have to be exact to not let food contact the glue. It was right on the bottles that titebond 2 is water resistant. 3 is completely waterproof.
Thanks for doing this! That very thorough, methodical approach reminds me of Project Farm tests. I have been using BondCrete which is a PVA concrete admix/sealer/surfacePrep here in Australia and getting better results than traditional wood glues, with the addition that I already have it on hand in 5L buckets, be interesting to know if you have something similar there you could test.
Bondcrete is basically just standard PVA, but in bulk. It certainly works well on timber, but surprisingly (for what it's sold as) it is not particularly waterproof.
Thanks Matthias. Very helpful. Would have been interesting to see a comparison with T-88 epoxy which is the go-to adhesive these days in wood aircraft construction using sitka spruce or douglas fir.
Just out of interest I'd love to see a dial indicator on that structural post to see how much it flexes when your doing the tests. Obviously wouldn't impact the result but its fascinating to see how much 'solid' things actually move!
Do you like making those graphs as much as you like performing the tests? They are always so nice, those graphs. It's definitely a highlight of the video.
This is a super helpful video. I had suspected that Gorilla wood glue was good, as I made some cutting boards with it a long time ago and they seem to have held up much better than other cutting boards I've made with different glue. Good to see some actual data to back up my anecdotal evidence.
Gotta love empirical evidence. I've used Titebond III for a long time and will probably continue since I'm very familiar with it's dry times for big glue-ups. However, for small, quick projects I've started using CA glue and it's a wonder.
I have a theory that the soft wood is more porous and absorbs glue hence the 5 minute epoxy beats the stronger slow epoxy. It’d be great to test some primed softwood to see if absorption is the issue
The polyurethane gorilla glue is completely waterproof, so it's a great choice for joints that will be submerged in water for long periods of time. It also works great on pressure treated wood that's still wet. I used it to glue together a wooden stand that's been submerged in water for the past four years. I finally got a chance to pull it out and inspect it, and it still looks great. (I'd like to strength test it, but I still need it and don't want to spend the time and money to build a replacement.)
every PU glue is waterproof - this is an inherent propoerty of polyurethane, 1-Component-PU-Glues harden by consuming water out of the air or the workpiece, water cannot reverse this reaction
Franklin (titebond) recommends full sanding of and smooth surfaces before gluing. They say rough wood will not bond as well. So looking at your softwood, I think that says a lot, vs the maple. The maple appeared to be smooth. Also wood that is not surfaced may have started to oxidize and therefore block the full bonding. epoxy likes a mechanical bond, so the softwood might be an advantage. Still an eyeopener. That opening test of hot glue had me on the edge of my seat. I never consider hot glue as a good perm bond.. so that first one was totally bizarre.
One suggestion is buy smaller containers of glue, label/date them, plus store them in your house, if your shop is not heated or cooled. Tape included. Thanks
The long set epoxy can take a week or more to completely cure at lower temperatures. You can mix some up and make a small form out of it and see that even in 24 hours time its still quite flexible and as time passes it gets stiffer and harder.
I would have loved to have seen some more serious epoxy than "5 minute" - some boat building stuff, 24 hour cure, - that stuff is incredibly strong in almost every scenario I've used it. I don't use it for everything, horses for courses, but it's excellent when you need it.
the foaming urethane gorilla glue works well on patching up crappy MDF stuff, its about the only advantage when a threaded insert or screw gets ripped out, Construction adhesive is junk, Jon loves it cause it works in his cold workshop and nothing he starts gets finished so there's no test to it
I long ago switched from TiteBond to Gorilla wood glue because most of my projects require staining. I always wondered whether there is a difference in strength because I never see anyone else using it. In my experience though, if TiteBond gets on the finished surface prior to staining, it is a huge pain to get that spot to take stain. On the contrary, with the white gorilla glue, it is much easier to scrape and sand any drips and squeeze-out. And, on lighter stains, if I miss a spot, the drips are a lot less noticeable because it does take some stain.
