It is bistable. They print it in a minimum energy configuration, then push to snap it into this configuration. Somehow the preload makes it possible for it to have constant energy over a range of deflections. I'm reading Reinier's thesis now to see if I can understand how that's possible.

​@@IainMcClatchieit's literally shown in the video that it isn't bistable

@@aonodensetsu Read the thesis. See how there is the funny S curve on the right hand side? The video doesn't show the configuration as printed.

@@IainMcClatchie this is not the well-known and similar looking bistable system, this is essentially a non-linear spring

I was getting all excited about this actual mature debate right here in the KZfaq comments, and then it just stopped 😢. I have no idea who is right yet!

with the plastic probably lower weight. with steel probably going to be a while. think of it like shocks on a car

​@@ARVashmy dads jeep shocks broke because he had like a ton of tools making ot very heavy (and the engine didn't like it lol) so i wonder how quickly it would stop working with all the weight in it

To varying degrees, plastics have “creep” which means it will permanently deform (take a set) when left under a constant load.

Generally that would have to be something a designer would need to account for. The material depending on how long it will be used for, will be decided. But as another commenter said, creep is the main thing you're talking and asking about

that is SUCH a cool and interesting question!

I'm sure you could do that with a trap door and a finger that pushes it open when the mechanism drops

can you explain a bit more of what you mean by resonant? I'm asking since you seem to know what you're talking about. or if its a bit lengthy of an explanation, can you point me in the right direction?

@@blakes8901 What I mean by that is that when he touches it, you can see it wobble up and down at a fixed rate, like about one per half second. That *I think* would mean that operating as a seismograph, it would respond dramatically to earth vibrations at around one per half second, and less to anything faster or slower. This makes it great at that vibrational speed but more "deaf" to anything outside it. It's not a deal breaker, it's just nice if your instrument is sensitive to a wide range of vibrations so you don't miss anything.

​@@NightRunner417thanks for the explanation! Every structure is going to have a resonant frequency, how do we construct detectors capable of intercepting a wide range of frequencies?

@@vishank7 I have no experience in this field, but i think a couple sensors with differing natural frequencies and some math between them would be the simplest way to solve the problem. It might get a little complicated with sensors being affected by each other's vibration though.

​@@vishank7 Hi! Sorry I took so long replying but it's super easy for me to miss seeing replies here. Ok, so... Let's build. First and most important in this rabbit hole of a thing is to recognize that it is in fact a rabbit hole. You're *never* going to go all the way unless you have the budget of NASA. Even then, they're never going *all* the way either, and I'll try to show you why. I've been chasing this myself for a long time. This is going to truly redefine "TLDR"... Apologies but it's how it is if you want to do real science in seismic sensing. First, you need to see seismic detection as a machine that includes the Earth itself as the key component. It thrusts upward and/or to the sides in typically microscopic amounts all the time, you just don't know it. Mostly this is on the order of less than say once per second but also can be MUCH longer. We want as much as we can get, right? Like from seconds per cycle to millions of cycles per second. SO, first, we need a *really strongly connected base* to the Earth itself. Concrete if you don't have bedrock. It's not overkill, it's necessity if you really want good sensing. You want anything soft like dirt or sand *out* of that equation because it will hurt your energy transfer. Now we can move on to the sensing machine itself, having coupled it as well as we can hope. MASS. Getting a huge range of sensing is primarily about mass. As the Earth shifts to the side or vertically, you want a sensing machine that says "Nope, I'm staying put!" That's what a big heavy mass does. Place a 50lb block of lead on a suspension and it's going to positively laugh at vibrations that try to move it. It can still be moved especially at very low frequencies, but by then we're talking ridiculously low, and pretty much NASA territory. It's fine to get us started in the right direction. SO, place that block in a very strong suspension like a steel spring spider. Block inside a steel frame with a four way horizontal connection to four very strong steel springs and those connecting to a very strong steel frame hard bolted into the bedrock or concrete. Keep everything as tight and small as you can to help reduce sensitivity loss. You can scale this up to increase low response, or you can scale it down to reduce expense and size but you WILL lose low frequency response. How you approach this is up to you. Now we need a way to milk the data out of that suspended mass. Here we're going to quickly realize that the magnet and coil idea, while cheap-n-easy is also cheap-n-nasty: It will ONLY sense so low in frequency. You'll miss everything that is too slow to get translated. As you beautifully pointed out, we live in a resonant universe, and magnet-coil setups suffer big resonances. We need something better, something perfect, something that is strictly logical precision tracking the position of that block. Optical is a great solution that doesn't sap energy from the block nor does it care about resonance at all. Hard-weld a metal "flag" to the block and use it as a beam interrupter for a defocused laser beam striking a large area photodiode. We're talking like a BPW34 photodiode with width of only a couple of millimeters, so we're still doing microscopic amounts of shift. If you wanted, you could use a much larger area solar cell and defocus the beam to be well over a centimeter, and then you can track large vibrations like happen in damaging earthquakes, but you'll lose resolution down to the very, very small things like footsteps, passing cars, someone dropping a soccer ball 100 feet away, whatever. Again it's up to you how ridiculous about this you want to be. Anyway, sensor and laser are mounted to the rock or concrete floor, and the flag you mounted to the mass block goes between them, cutting HALF the light of the beam. Now you just attach the sensor output to an ADC and graph the data in real time, and you just gave yourself a whole new set of problems like temperature drift, sag in the springs over time, and intruding ambient light. You MUST shield this against outside light, preferably inside a box with a light absorbing inner coating to guard against beam scatter interference. But hot damn will this setup give you low frequency response and incredibly high sensitivity. You can do all three vectors of motion on three laser/sensor setups and detect motion on a truly microscopic scale, just have to accept that at some point you're even going to be fighting against tiny air currents. But it's pretty brilliant, and you'll be astonished at how far away you can stomp on the floor and pick up your own human activity. AND you get a fantastically wide range from tenths of a Hz to god knows how many kilohertz. That upper limit is going to become an ADC limit problem rather than the design itself. My own micro version uses a totally different approach to sensing and I just swallow the fact that it won't detect slower than 1 cycle per second. I figure it's good enough and it somehow manages to easily hit everything from 1Hz to many KHz with sensitivity so strong that I can capture myself stomping lightly on the floor anywhere in the house while the device sits on my concrete back porch floor over 30 feet away. All this from a little 3d printed plastic box about 5" square. 🙂 The design is still good enough to show me real earthquake energy like from mag 3 ish quakes but I wouldn't trust it to not max out past that. Anything stronger and I'd be no doubt seeing square waves in the output till it settles down. I'm more interested in micro-vibrations anyway. When the kid two houses down is running around in their backyard like a maniac, I want to see every footstep thumping that realtime graph. When a car passes down the main road 200' away, I want to see those tires shaking the landscape. I find it fascinating. Anyway, I hope all that helps. Seismic science is a truly fascinating thing, and far more challenging to dig deep into than one might at first think. It makes me think that giant machine like LIGO would make for a truly amazing earth vibration sensor, down to vibrations that are barely there.

Yes, if you are referring to a spring in the Hookian sense (linear relation between force displacement and constant stiffess). This mechanism has non-linear force displacement and different stiffness along the displacement. It starts off with high stiffness, followed by a near zero stiffnes, and then high stiffness again. It is because of the near zero-stiffness (but positive force, due to the stiffness at the beginning) that we refer to it as constant force. Since gravity acts as a constant force, it can be used to compensate gravitational forces. Gravity compensation can be used in robotics to reduce motor torque for example.

there is more fancyness in it, than it sounds!

@@ReinierKuppens I made this kind of mechanism using monolithic aluminum. It works better than 3d printed plastic. I can tune the thickness of the buckling spring using sandpaper to get almost zero stiffness. But the buckling spring is really sensitive to overstress. Really glad to see your mechanism work well there.

@@nikkiharpana8621 Def would be with aluminum. Its cyclic properties are shit.

"The primary advantage of a gravity compensation compliant mechanism over a traditional spring is its ability to provide a CONSTANT FORCE over a range of motion. This constant force capability is crucial for effectively counteracting steady forces like gravity, which remain the same regardless of position."

not so much, Best buddy designs the cameras and rigs for the cars they shoot movies with. you dont want any material that flexes.

I was wondering the same thing of "Could that make a stabilized camera mount"? at the end.

I think it's a scale that can only measure one weight, because the force is constant (unlike regular scales which have a spring in them that gives a bigger force with more deflection)

Correct, the design is monolithic and 3D printed. The stl file can be downloaded here: www.thingiverse.com/thing:5462230

​@@ReinierKuppens How did you 3d print this? FDM? What material are you using? So cool!

@@gedr7664 last I heard he used PLA

It has extremely non-linear stiffness - very stiff in the beginning of travel, very soft in the middle, very stiff at the end. It allows near-constant force in the middle of the range of travel in a compact package.

This is the first comment I understood.

It is not explained in the video but a constant force mechanism is quite unlike a spring but closer to a weighing scale. For a spring, hookes law applies, so the force is not constant but linearly proportional with the deflection. If this mechanism behaved as a spring, the bucket would go down a distance proportional to the load applied. Instead the force here is constant which you can think of there being a counterweight on a balance. One side of the balance can be loaded and the other side has a fixed weight of about 1.2 kg. When you load the scales with less than 1.2kg, the unloaded side stays up, more than 1.2 kg it goes down and close to 1.2 kg it is balanced and floated. Around the 1.2 kg the effective stiffness of the mechanism drops to zero.

​@@lkwakernaak I can see that this mechanism introduces some nonlinearity, but it is far from constant force. Constant force mechanism would let the mass float in a stable manner, within a wide range of different positions. In other words the mass would remain still at the level where you left it. This is not the case here. When you swing it, it goes up and down a couple of times and later it always stops at the same position, meaning that below the force is higher, and above the force is lower. Like in a spring, nonlinear or not.

@@jakubp1790 I agree that it not a perfectly constant force but I do think it is pretty close. I wouldn't classify this as a spring as it doens't behave like a spring around the undeformed "up" configuration. It is behaves like a very weak spring around the middle position that it reaches at a force of 1.2 kg. Perhaps this mechanism could be tuned to be closer to zero stiffness around a larger regime but it is either going to be a very small positive stiffness like here, or a negative one at wich point it is going to be bistable. I believe there is some plasticity showing up here as it first stays in the "up" position before Reinier moves it down and then it won't stick in the "up" position anymore. Meaning that there is probably some room at least to tune the mechanism a bit more.

easy measuring of drug portions

its interesting

expense and density?