A great way to start the weekend: with a new video of Matthias Wandel 😊 I‘d like to see more cooperation projects - especially with Marius Hornberger 😊
@bengy5959 no, it's not rude to correct spelling mistakes. The common view that it is really needs to stop. It's never bad to correct someone that is wrong, and your only reaction to it should be to learn from it.
This testing continues to convince me that modern wood glues are so strong that with any sane joint design, the failure is going to be elsewhere, or the wood coming into the joint, not the interface between the pieces of wood.
The plywood ones had a radius inside corners which adds strength And if the nuts were super glued in that would add strength to the main beam. You too create some of the best and most educational videos
That's the thing about testing: you sometimes/often find modes of failure that you weren't expecting. You redesign and test again, and again ... until you are satisfied. I'll call this a successful test.
Because you did not apply the edge clamping load to the spine it only sees bending and the direct tension from the pull. The clamping load would on average increase stresses, however, may reduce extreme fiber stress because it’s in the opposite direction. This would make your results more conservative than reality so no great loss. It’s strong enough and your test good enough. Just some food for thought.
This is what's been on my mind as well. I see a few other issues as well with this test rig setup. I wish a bit more focus went into the execution of this research. We all know Matthias is perfectly capable of something better :)
For comparison purposes, it works just fine. However, I agree that the physics and associated math is much more complex. I believe Matthias is well aware of the potential shortcomings of the setups if one were wanting truly accurate measurements.
A few of these weren’t technically a “joint” failure but either a failure of the timber itself or the pins simply pulling through where they’re drilled. Would be good to see a pin holding the inside of the joint vs a drilled hole.
Hi, Mathis The question has to be asked What sort of Pressure Are they likely to on, They’re only keeping some edge banding on by what you said so pressure of Say 20 pounds is maximum they're going to have so all of them will work
if the center hole weakens the frame so much, I don't understand why you can't just make that side thicker? why is this overall thickness of the clamp so important? or just add a metal strong tie on each side, they are cheap enough.
I'm no engineer, but are you sure you can double the result from your force vector calculation? The failure of the frame is already force being applied to both sides, and you're just translating it to a different (symmetrical) vector applied to the same part. It would seem to me that half of the failure force would be applied by each cam, and thus you can't double it.
Ah I see. Marius Hornberger's edge clamp video is a Patreon exclusive. Was wondering why I couldn't see it on his channel. Edit: Nvm. He just uploaded the video.
the quick convo in german got me curious. turns out matthias's family moved from southern germany in the 1980s, so his native language is actually german!
When you made the plywood versions you had the grain oriented 90 degrees from the original plywood versions. You would have to count how many plies were passing by the center hole (perpendicular to the hole) in your version compared to his version to see if this explains the difference in strength. The plies with grain perpendicular to the direction of stress don't contribute to strength and are ignored in actual design of plywood. You can find books listing the effective area and bending properties for normal construction plywood considering the plies in the direction of stress, I'm not sure if such a thing exists for baltic birch with its different number of plies.
Here, as in most mechatronic systems, it seems that stiffness is a better metric than ultimate strength to say "which is better". Nice! Congrats to Marius for being 'friended' on YT by Matthias. It's official now :)
Does depend on the system, more than a few things are very reliant upon the elastic flexibility for function too. And sometimes you want the failure mode to be non-destructive and give you warning the mechanism is overloaded, which doesn't tend to be the case for stiff structures, if its too stiff the tendency is to work perfectly till it fails catastrophically.
from my understanding of clamps much of the force comes from the elasticity of the frame material and not from the elasticity of the threaded rod for example so stiffer frame doesnt necessarily give you more holding force for a given torque
Great video Matthias! Another interesting idea may be to test the clamping force over time. If you can for example clamp to within say, 20% of force before breaking, then leave it overnight and plot the force over time for several hours. I'm not sure exactly what this would test, or if clamping force would even go down over time, but it would be interesting!
I'm no structural engineer, but wouldn't the test of the glued joints' strength be more appropriate if the two legs of the sample pieces were compressed together in a vise like test structure? This would eliminate the wood failures due to the drilling of holes in the sample pieces or other abnormalities.
I am more then surprised, that the 2x4 (?) on top with the notch sawed in (whith sharp edges, which are way weaker then rounded ones for example) survived all this.
Very interesting! I wonder if you could find a way to measure the deflection of the joints as you apply the tension. I’m interested in seeing the force-deflection curve for the joints.
Forgive me if the answer I seek has already been delivered in the video ... BUT how much force would really be needed for each clamp to function well in the mode of use intended?
The plywood were ultimately much stronger than the other ones when there is no hole drilled. I think the best solution isn't to go with a mortis and tenon, but to simply reinforce the plywood where the holes are. That would be easiest to mass produce and the strongest.
could you increase the leverage? make the arms longer, that way you put more force on the joints whilst the rest should get lett right? or am i mistaken?
I think it would be a cool video to throw 3D printed parts into the mix of this test. But I do realize this is more of a woodworking channel but would still be cool. Great videos!
My best guess is that 3D printed parts aren't as strong, but that would depend on the printing material and the printing technique..so a whole other story
Love these failure test videos! I noticed something near the end with your math though - if the reasonable max load across the arms is 150kg, why did you double it to get the perpendicular max load? To me that doubling was already taken into account naturally by your test since it was loading both arms at once.
Two scoops, I think the load is doubled to account for the mechanical advantage of the lever bar across the top of the test unit, the force is applied twice as far from the fulcrum as the load point. In use the clamp will always be a single unit because the load will be balanced between the two legs. If the spine of the clamp were secured and the load would bear against only one leg, the total would be doubled, but in actual use this does not occur.
@@harlanbarnhart4656 This totally would make sense about the test setup! I went back to that part of the video to see if I missed him explaining this, but actually he says "and since we're pushing from both sides" in order to explain the doubling of the force. So I'm back to being uncertain of the result.
When you clamp on something, the load tries to bend the wood, but because of where the load it applied (Pin through hole) it instead split the fibers. ( This is where larger and "weaker" grade 2 bolts and lags should get to outperform structural screws rated higher that end up splitting the wood, That would make a good theory to test too!) If the arms were built like a torsion beam/box, with larger holes with a metal pipe spreading the load to a washer or insert, the split might be avoided.
I can't find Marius' video despite being a long time subscriber. I think I saw it pop out on my Subscription page, but now it disappeared. Also going to his channel show the latest video being the one about the pantry!? Am I missing something here?
@@matthiaswandel @MariusHornberger No worries guys! I was just confused because I saw something pop up and then couldn't find it anymore and then Matthias referenced it... Anyway half way through Marius' video now and appreciate both your videos as always. Thanks for the excellent content!
I like his clamps but for the small pressures being put on the treaded rod he should have used a much smaller diameter rod to strengthen the body of the clamp
What was the species of wood for the clamps themselves? They all failed as I expected (the material, not the joinery). While plywood would be fast/easy I wonder if he just laminated the horizontal clamp members, from 2-3 layers of solid beech, each at .125" thick. These would look nicer than the ply.... Interesting for sure.
since yield is an important consideration for these clamps instead of just UTS, would it be possible to add actual stress/strain measurement and graphing to the stand?
Seems like a lot of time could be saved if your collabs started out with design specs. That is what interested me a lot about the process: I actually have no idea how much force you need to edge band something. most of my experience is with super low clamping force hacks like masking tape clamping, putting some elastics around one of those dollar store spring clamps, and the like. Which obviously generate nothing like the forces you guys are clearly targeting. Are those forces needed? what applications are they needed IN? I have mostly seen edge banding as decorative, and if you need strength, you clamp both sides of your panel glue up. (with something like a bar clamp on an assembly table). Is this for stair nosings or something? Most stairs I have seen, the treads are assembled (nosings and all) before install, and the joints are both mechanical and glue. I found your force diagram at the end to be the most useful, as spoiler alert, none of this ends up mattering at all as the failure mode of Marius's clamps is his rubber friction pads failing, at nowhere near the design strength of the clamp arm joinery. If the individual components had been tested individually, I think it's safe to say a ton of build time could have been saved just by saying "none of this matters because our rubber is going to tear long before we reach these test strengths". Neat hack with the wood wedge jammed into the gap in your screw clamps to apply force. that looked strong and obviously, takes up basically zero storage space. I will definitely try to steal that if I ever figure out why someone would need 100kg of clamping force for edge banding.
