 So, hallo zu meinem Vortrag Designing for Laser Cutting. Ich bin vorhin Festi, ich bin der Autor von Boxes.py, ein Boxgenerator. Ich mache Laser Cutting seit jetzt fast zehn Jahren und ich würde sagen, ich habe mir noch sehr, sehr viele Dinge da nicht gemacht, aber ich dachte, ich mache mal jetzt das gute Zeitpunkt, um mal ein bisschen Fazit zu machen und zu gucken, was man so gelandt hat. Bombs Laser Cutting ist interessant. Sie sind im Wesentlichen schnell, relativ einfach zu bedienen, meiner Meinung nach die einfachste Maschine, die wir im Space haben. Sie sind präzise, also wir machen so Fingerzinken auf Plus, Minus, immer so in 200 mm Schritten, um die Pressung einzustellen. Nachteil ist, sie sind eben halt Inherenz 2D, das ist einer der Gründe, warum sie leicht zu bedienen sind. Und das heißt, man muss sich irgendwie das überlegen, wie man von den 2D dann zu 3D kommt. Und das heißt, außerdem, meistens, wenn man was ausschaut, muss man so nachzusammenbauen, weil die Teile eben flach sind. Worüber wollen wir jetzt reden? Der Vortrag ist im Wesentlichen so zweigeteilt. Ich will im ersten Teil so ein bisschen über Engineering. Ich wollte eigentlich auf Englisch reden, oder? Warum rede ich Deutsch? Das ist fivernt. So, der Talk hat 2 Parts. 1. Part ist about Engineering. So, we are talking about Strength and Stiffness, which are one of the major properties. We will talk a bit about Blyworth as a building material, which has special properties that are distinct from many other isotropic materials. And after that, there will be a second part, where we talk about solutions to common problems and what features we can use and how to use them to their best abilities. Short disclaimer, I'm not a mechanical engineer. The talk will not make you one. And so, please don't use this to build something that kills people. If you want to build bridges or something, please study first. So, the first thing that might be of interest is that the things we build are strong enough to not break. There are a couple of different properties that materials or things have, and they can be pretty confusing. We are not going over every. We are only looking at two of them. There are a couple of more, like toughness and hardness, which are related, but not the same. So, Strength is the ability of a piece to withstand forces without breaking. There are basically two things. One is the strength of a thing, and the other is strength for material. Material has strength, which is basically a pressure that they can withstand. And they can withstand it in basically three things, in compression, in tension, in inschirring. And the thing itself can also be bent or torqued. And the stress they can withstand is a pressure, so it's measured in Pascal, which is Newton's per square meter. When you design something, you need to basically look what strength your loads would result in, and then look if your material is strong enough or adjust. And you often want to put a safety factor in that, so you want to make your thing three times stronger than it needs to be. One reason for this is real safety, like in case you miscalculated something. And this other thing is fatigue. A lot of Materials will get damaged over time if they're loaded repeatedly. And so sometimes three is a good number in general, but different applications have basically different ideas on what safety margin is necessary. Or can be afforded a lot of applications like aerospace engineering or Aerospace, which are tight on weight, might choose smaller values. Something that sounds similar, but is different is stiffness. Stiffness is not about breaking, it is about the give a part has, if it is loaded with a given strength. So the general thing is everything is a spring, so whenever there is a force, something moves. Often we can't see that because it's so little, but it does move. The material property that's affecting this is the Young's modulus, or Immodule in German. It's a material property, so basically depending on what material you use, your part will be more or less stiff. The second thing that's important here is the second moment of area, which we will go into in detail. Don't be too shocked. We will be talking about a few formulas, but they are not for you to actually do calculate small, getting a feel of how this all works. The main takeaway here is that stiffness is often more important than strength, because you will often not load your part up to breaking forces, but often, even if things shift around too much, that's no good. So mechanism may no longer work, your gears might be misaligned, it might just feel not great if you touch something and it's squishy, you often want a certain stiffness, even if it won't break. If you don't need any stiffness, you can use fabric, which has no stiffness at all, and it won't break on you, but you can't stand it up. So loads and stresses, when you really think of a thing, there's stresses or forces that a minuscule piece of your part can experience. So these are basically fundamental on the Minecraft cube forces. And there are three of them. One is compression, one is tension, and they are shearing. Shearing is if you basically try to move things past each other. This is often only seen if things actually fail in that way. And it's called sharing for a reason, that's what shears do. So this is typically cutting, but if you have like rebar cutters, they will shear the steel by just pushing it to the side. And there are loads, which basically means forces you put in a part, and that will lead to stresses, but they're not fundamental. One of them is bending, which can lead to all three of the stresses above. And then there's torsion, which means basically twisting apart from both ends. So there's a whole, for bending alone, there's a whole theory. And we will not go into that in detail, but just know there's a lot of more mathematics, and there's a couple of formulas for the easy parts, and there's finite element analysis for complicated parts, but typically we don't need that. We just want to get a quick idea of how this all works. So when we're bending something, there's a bending torque. A torque is basically a force multiplied by a lever, and all the bending forces are like this. And the length of the lever depends on the scenario. We'll look at two very simple, the two simplest ones to get an idea of how this feels. If you don't like bending, you can use a truss. A truss is a construction with beams that are all loaded only in compression or tension. So there are none element is experienced at the bending forces, but then there's trust theory, which tells you how all these forces work, and that's another kettle of fish we are not going into. So here's the simplest thing, that's a cantilever beam, that's basically just fixed on one end, and there's a force on the other end. It's basically the simplest thing, it's just a lever. And this is basically the definition of torque. The torque moment is the length of the beam times the force applied. So, that's all that's this. There's the formula for how much it bends. So, it's the force, the length is cubed, so, that's the reason why long things are very floppy. So, why you can, if you've ever seen a railroad track, that's not has been bolted down yet, they flop around like noodles, although they're like still in this size. This is one of the reasons, because the L is cubed, so they are very floppy, if they're long enough. And there are two other letters that are important. One is E, this is the Young's modulus, that's basically the stiffness of the material. And the other is I, this is the second moment of inertia, that's basically the area of the beam resisting. There's a formula for this, which we'll look later, but it's basically how the beam looks, in what the shape is and how, which size it has. And the bigger it is, obviously the less floppy it is. So, there's a lot of other formulas for other easy cases, like having the load distributed differently, a typical case is having basically the weight of the beam, looking at the weight of the beam, basically the weight being distributed over the whole length and stuff. That's something you look up, if you're interested. We are only interested in saying, well, this is a beam, there's a torque, that's it. Here's another case, there's a beam, there's a load, the difference is now it's supported in both sides. The important thing here is that they're, while they are supported, they are not hold in place, so that when the beam bends, it can twist in the ends without anything resisting. Because if you do that, the stairs changes a lot. So, we can see the bending moment underneath, the bending moment is biggest in the middle, and the maximum moment is only a fourth of that, of the other case. Why is this? It's very simple. If we hold the beam here in the middle, and then we have two forces that pushing it up, and this one in the middle, if we hold onto here, we don't need any force. So, it's basically like, this is a beam that's held in place here, and then pushed up here. So, we have a beam of half the length, but we also splitting up the force into two sides. So, that's the reason why it's only a fourth of the other case, because it's basically the same case, half the size, half the force. And as such, the bending is also a lot less. It's basically the same formula, but instead of dividing by eight, we're now dividing by 48, because of all the smaller forces and lengths. So, what do we want to keep here? There are different types of these bending scenarios. They are basically different by the factor in front, and otherwise, they are basically all the same. It's basically all the lengths, and the lengths differ depending on how the beam is put in place. So, next thing we are interested in is how this second moment of area, or inertia, that's used interchangeably, looks like. And I have brought here a bit more complicated example, but it's basically rectangular cube, with a hole in the middle, that's also rectangular. And basically, we are just subtracting the term from the outside rectangle. We take the box from the outside rectangle and subtract the one in the middle. And the interesting part is that the moment is basically proportional to the width, B, and the height cubed. So, that means, if you're making this twice as height, we get eight times the resistance. So, this makes a huge difference. So, whenever we have a construction that we think looks a bit flimsy, increasing the height will do a lot for us. Like doubling the height will basically make a world of difference. And so, often, even increasing a little bit will be sufficient. This is also, of course true, if you have other shapes, they have other factors in front, but that doesn't really matter. So, for example, if you have a screw and think, well, it looks a bit flimsy, let's take an M8 instead of an M6, that doesn't sound like a huge difference. Yes, it is a huge difference, because the width is also increasing, so it's basically to the fourth power. Basically, doubling the diameter of an axle increases the moment by 16. So, going from M6 to M12 is a huge difference. That's maybe something for all those driving contraptions that are around here. So, if you, there's a lot you can do with just the diameter. So, on the, not quite as good news as what's the next, so there are two, we are still looking at two things. One is the stiffness, and the other one is the strength. And the strength, for the strength, the question is what's the maximal stress that the part can withstand. And the part will fail at that point, that is mostly, that's loaded the most, and the most stress is at the top, very top and very bottom of the bending beam, of course. And the stress throughout is basically the moment, the bending moment divided through the moment of inertia, but it's multiplied by the distance from the center line. So, we're losing one power. So, basically, while this is called Z, it's basically proportional to the height. It's basically the height, the half of the height, depending what shape we have. Some shapes have different, are not symmetrical, so it may be a little bit different, but basically this removes one power for the strength. So, the strength is one power less than the stiffness, but still it's still, at least, quite tragic with the height. So, it's worth increasing the height. So, there's two more things we want to look in here before we move out of this whole theory stuff. Just torsion, which means twisting a piece around from a mathematics point of view, that basically works very much the same as everything else. There's a polar moment of area that we are using. It's also power of four, because we are, of course, basically we have, last time we had power of three of the height, and then the width. Now, if you're looking at a circle, for example, with the power of four of the diameter, the problem here is so far we could basically choose which axis we are wanting to bend to. And if the other one is weak, it doesn't matter that much. So, if you have something that's very high and slim, yes, it's not as great as being wider, but it's only linear. Here, things are different, because we are rotating it. So, it's basically the formula for complicated, more complicated shapes are more complicated than just bending. And basically, we want to have as much of a diameter everywhere. So, having a circle is obviously the best shape. Having a square is still good. Having a very slim beam is not great. And the worst things are slotted things. So, if you have a pipe and cut it in length, it can't resist much, because basically, the surface that has to carry all those force has to basically take a 180 twice, and that's really, really bad. What happens, if you actually do that, is that the two sides light against each other, when you bend those. And if it's connected, instead, if you have a tube, basically, they can't slide anywhere. And the forces are going basically as a spiral from one side to the other, both in tension and compression. So, that's something to keep in mind if you're looking at torsion. Another fun thing is buckling. This happens whenever a member in compression tries to evade the force by moving to the side. That's bad, of course, because if you're compressing something, you're basically counting on the area being there, and as soon as it buckles, so it's bending upwards, and then the bowing makes the geometry less stable, of course. So, you're turning compression into a bending moment that can lead to catastrophic failure, if it basically collapses on itself. That's one reason why we can't make these beams as thin as we want to. For one, it makes the torsion stiffness very weak, but it also makes it prone to buckling. That's why a lot of construction beams are using typically IRH-Beams, which are wider on the top and front. And they could, of course, just have a round bar on the top and front to have the area, but making them wide, for one, gives it some stiffness for bending in the other axis, but also prevents buckling. So, having something too slimm is optimizing in one direction, but keeps you weak in others. One thing is, there's no buckling in tension. Whenever you pull something, it will pull straight. So, that's one thing to avoid. And some constructions are using these cranes, for example, that have the tension member only as a wire or a steel bar, which is lighter and doesn't need any additional width to get stiffness. And they have the width only in the compression part of the construction. As we are in bending, there's another fun thing that I ran into. I've not found an English word for this. It's called Schubladen-Effekt. Basically, if you have a drawer that's too wide and too shallow, it will bind up and it binds up, because the lever, if you're pushing there, the leverage for binding up is better than the force for pushing it in. It's something that's used, for example, in F-Clamps, where you screw it in and it locks in place, but if you're losing them, you can move them up and down. And that's also something, if you do constructions where you have sliding things, make sure that the length of your support is long enough to deal with the offset of your forces. So, I did make this nice little clamp for clamping material to the knife table of the laser cutter. The first one was too short. The sliding part was too short and would just not move at all. As soon as you're not, unless you're pulling it fully straight and just making it like a centimeter longer, fix the problem completely. It's really amazing how much difference there is being one side of the border or the other. Of course, you can lower friction, that also helps, because that's basically friction against the ratio here, but often you're limited or have already done everything to do that. Good, that's the theoretical part. Next thing to think about is the material we're using. I'm talking plywood here. The plastics are a bit different, but there are not that many easy-to-cut plastics that can be used for construction with laser cutters. Probably sterren, probably, there are a couple more, but they are all not cutting that great. Acrylic that cuts really well is not a great construction material, because it's relatively brittle, which is another material property we're not talking about today, but it makes stuff pretty easily. Wood is a lot lighter than steel, which is another usual building material and something that you will encounter when putting other stuff onto your laser-cuts. Things like screws or ball-bearings or whatever. The main wood is an actually pretty good construction material. It is relatively strong, giving it its weight. The problem is it has the strength only in one direction, and it's very weak in the other direction. This is compensated to some degree by plywood that are glued in two directions, so you have at least half of the strength within the sheet, but you still have the weakness across the sheet. So if you glue the sheet to something, you still only have the weak direction of fiber to fiber to hold it in place. As a result, you need 10 to 20 times the length to get the force into the sheet. Over the length, basically the force can move into the depth of the material. This is, to some extent, better, if you have plywood with a lot of layers, because basically each layer can do this on its own. Not if you're gluing only the top layer, but if you're putting the forces on all sides, basically moving them around. For example, if you want to move the force around the corner within one piece of plywood, you basically only need to move it from one layer to another, which may be only a millimeter, so you only need like 10 millimeters or 20, so you only need a centimeter to actually be able to move this force around the corner. Which is still not nothing, if you're thinking about all the small figure joints we're using, but it's much more doable, if this was a massive piece of wood, that we have to glue to something else, where we would need much more area. So, the next problem is that steel is much harder and much more heavy than wood, that makes it also hard to move forces from a steel part to a wooden part, because the wood doesn't have the strength to actually move the force somewhere. So, you're limited basically by the strength of the wood to take the forces from a steel part. So, the obvious way to do that is to increase the area by using washers or using oversize your steel parts, use an axle that's thicker than it needs to be, so you have at least some area to hold on to. But often that's not desirable, because for one it gets more expensive and then it gets heavier, and if it needs to be moved around, that's bad. One solution can be using dowel bearings. They are like LEDs, they are just too cheap, if they are small and everyone uses them, and I'm not a big fan, but they have their uses and they can be used here to just increase their area, because the outside of the ball bearings is of course much bigger than the inside, and you can still use a thinner axle with that. Another option is to just not use steel. So, basically try to laser cut the pieces you need also from wood or some other material, or use something that's lighter, like some plastic parts. I've got a couple of numbers here, maybe I will not really go over them. The things I always keep in my head is the density of the materials, which is for wood is like 0.3 to 0.8, depending on what you have, populars or relied. Birch is more on the upper side, there's hardwoods that are even heavier. Aluminium is 2.7, these are all kilograms per liter, or tons per cubic meters, if you're building something bigger. Steel is about 8, so that's a huge difference, basically a factor of 10, as I said before, even for heavier woods. We are not going over all the other strength. You can see that steel especially has a huge margin on what strength it can have. Mild steel is not really better than wood, when we're talking about specific strength, so if we count on weight, there are hardened steels that are very, very good, but they are also in pain to work with, so they don't really compare in the sense that you can easily less cut them and put them together, they need heat treatment and all kinds of other stuff. Metallurgy is a black art, we're not going into that. So, what are the tips we can learn from that? One thing is try to keep all your forces within the walls, because the height of your wall is the height of your beam and it is multiplied, or it is taken to the third and fourth power, so second and third power for strength and stiffness, and that's a lot, even for smallish things. If you're looking away from just a single beam, but for larger constructions, it's good to have things enclosed on all sides, basically. We have a box like this, this is relatively strong, but this is for example weak in this direction, because here we're basically relying on the bending of the joints here, so if we had a wall in here, this would be much, much, much stronger. So, often it's worth just putting a wall somewhere and even if you cut a huge hole in them, it still stiffens out things a lot. If it is round, you basically have a triangle-shaped quarter-circles in the corners that will stiffen those out a lot. If you have a larger contraption, there may be parts that you really care about, that they stay in place, and what I would suggest is you basically in your mind grab them and push them around and look for, if you have support for all six dimensions or actually 12 basically moving in this axis, this axis, this axis and then also twisting in all three directions. Und if you have all of them properly supported, it's very surprising how stiff and how stable something can be. I have built a cocktail robot and there's a platform hovering basically in free air and all it compresses is very flimsy, three millimeter arcs, there are six of them, two on each side and one going to the back. And although you can basically bend them with your small fingertips to the side, in combination all those arcs are supporting the platform very, very well and it doesn't move barely at all. Even there's some arm on top that moves quite quickly. So that's the next tip, basically make sure that the stiffness doesn't come from one part that's strong, but from the overall construction, make sure you support all possible forces. And the next thing I would suggest, although we have seen all those formulas, I would rather suggest not to use those. The main problem is you basically have to set up a scenario that you then do the calculation on. So if you need my talk for doing the calculation, you will have even more trouble setting up a proper scenario. So you need to make all kind of assumptions on what are the forces, which are the directions of the forces, which forces I care about. And that's rather troublesome to get that right. Because of these huge powers of these strengths, eyeballing, that works really, really well. And if you're in doubt just at 50%, and then it typically will work out fine. So, that's the theory. Let's look into some of the techniques that can be used with laser cutting. And of course, the first question was, how do we get from 2D to 3D? And the simplest way of doing this is just gluing layers on top of each other. I, for some reason, have not found this box, that should have been in our Hacker space somewhere. So what can you do with that? You can do lips and slots to slot something else in there. You can put mounting features. So if you want to have raised points where you can screw in PCBs or stuff like this, that's all the things you can just glue on top of each other. PCB and Glowshoes, if they are flat, it's probably... I'm a Sweebian, so I'm hesitant of wasting material, but you're still for very thin, low-profile stuff. It's probably easier to just glue up four layers of stuff on top of each other than construct a more complicated thing. This example on the left is a bionet lock for the lid. So basically you put it in and twist it, and then it locks in place. Another way of doing this is just to use a 3D model and put it in a slicer. That's how this thing here got created. That's basically just a 3D model cut up with a... Ich denke, es ist sogar ein Standard 3D-Printing-Slicer, das hat eine Option, um nicht nur den Path zu haben, sondern nur die verschiedenen Läden als Geometrie zu haben. Ein Tipp für das ist, es ist oft nassig, um das in sich zu nähen. Wenn man so etwas hat, ist es wahrscheinlich nicht wirklich ein Tipp, aber man kann es in hier ein Tipp geben, weil man nur die gewisse Gewichtung für die Waffe braucht. Man kann es in verschiedenen Dimensionen in unterschiedlichen Dimensionen in sich zu einem anderen hinlegen. Das ist auch etwas, was hier gemacht wurde. Die inneren Läden sind von der gleichen Läde, nur eine Läde, und die inneren Läden sind einfach geklappt. Ich bin nicht ein großer Fan der Kurve des Lasers, als die Maßnahme, die in der Part zeigt. Aber für das ist es okay, in meiner Meinung. So, die Frage ist, wie du das sammeltest? Wir sehen hier ein Lippe auf dem Boden. Mein Lieblingsmethode ist, nur zu benutzen, Locating Pins, das ist etwas, das in Maschinenkonstruktion gemacht ist. Aber hier hast du nur kleine Halsen, 1 mm ist gut, nur kleine Halsen, um die Piste zu locieren, dann kühlen sie, und wenn du, in den meisten cases, die Halsen rauskühlen. Wenn du etwas dicker hast, kannst du auch nur die Halsen inhalten und sie inhalten, das ist ein Reinforz. Das ist in meiner Meinung die easieste Weise. Ein weiterer Weg ist, zu einfach die Lasercut-Packs zu benutzen, basically have a square hole with one material thickness squared, and then cut small pieces either with a top hat, that locks them, that puts them on the right height or just flush and hammer them in. Another way is to have C-Clamps, basically have a slot on the side and have a C-shaped piece, where the layers are all basically stacked inside, and if you have two of them, they are also located. For stuff that needs a lot of torque, like small gears or pinions, you can also think about putting thicker metal rods, like thicker nails in there to transfer the torque, that can be useful. Finger joints, I guess everyone has seen them already, they are everywhere, so I'm not talking about the obvious stuff too much. You can use them in other angles than 90 degrees, but they lose a lot of glue surface area. So they are much, much weaker, they look cool, but don't rely too much on these kind of finger joints. Also it's not possible to design them in a way so that they create closed surfaces. So we'll always have holes or gaps in between. You can get rid of the gaps on one side, but you can get both sides without gaps, even if you're willing to send off the overhangs. Generally, the smaller the fingers, the stronger the joint, because it increases the glue area, basically gives additional glue area on top and bottom of the fingers. This is useful, because bending like this is the main weakness of these kinds of joints, because that's also bending. We basically have the whole length of one side as a lever to pry them apart. We have the material thickness to withstand that. So, that's a good reason to always have a top and bottom piece. That's one of the reasons why there's this round part on top. It keeps all those together and prevents from bending stresses getting onto those non-90-degrees joints. There are two ways to do these finger joints. One is symmetrical, that's what we are doing here. So we have two sides, basically one with the fingers going out and one with the fingers going in. There's a different way of doing them, which are not shown here, which is basically they're symmetrical, where all the fingers are on the left side and they're going in on the right side. Basically, if you have a part and they have a direction, putting the pieces together, basically with the same side up, they will always fit, because this goes this way and this goes this way, this round. So they will always fit, if you put one on top, it will also work. The technically much better version is using finger holes, that's basically fingers on one side and holes on the other side. That's our arcade, that uses this. For one, it increases the gluing area, so now we are pressing always against material. So we are always, we basically increase the pressure when we are bending it, so it making it harder to pull out. The thing that's a bit tricky is if you want to put two things from two sides at the same place, that can be done, but you basically have to skip half of the fingers on both sides, which can be a bit tricky, whether they are, whether they have an even number of fingers or not. Often it's easier to avoid this and just offset the two walls and have them at different levels. Maybe even one or two material thicknesses apart. So they are basically material in between to connect all the forces. Another way, if you want to cross over walls is to use slots, they are not very stable, slots just means you have slots like this and put the materials together like this. This works really well for one direction, because the thing slotted from above can be fixed to the bottom and so you have a continuous plate on top in the bottom that could flop around, so that you can go back. The problem is the other way around, that those are basically unsupported on top. So if it's upside down, if it's on the side, like for some storage, you make sure that those are the ones that are vertical, so that they can take the load in the right direction. If you put them this way, they will just bend down and be not supported well. You can of course join them in front, with some other piece that connects both directions again. If that's something that's useful or feasible. One thing I really like, but it's actually not that useful, are flat dovetails. The only really serious use case are those rounded boxes, where we have one piece that wraps around and we need to basically join them back in front. They are a good way to create pieces that are bigger than your laser cutter. These are pretty big ones, but they can be made very small, like only 3mm or something. And then if you put them with a lot of pressure, with a lot of, if you make the cut very tight and you combine them with a hammer, they are basically like one piece. They are surprisingly strong. Another thing that just looks fancy is flex. I've done quite a lot with that. Flex works well, if it is constrained to the radius, if you have constrained the maximum radius, you can bend them. So there's a lot of things that use those as hinges and I've never found them that particularly well working. Because of fatigue they will always break at some point. Also, if the problem here also is that the forces in this direction are also created by these very tiny pieces and basically there's very little wood fibers that go in this direction. And if there's some defect in the plywood at exactly that point it will just break. So, it's very fragile. Also, also, in this direction there's no way it's bent too much. It also works reasonably well if you do it in corners, like we saw before, but it's not that great if you allow users to over bend it. That's something you just need to avoid. Hinge hardware, something that can be done with laser cutting works reasonably well. I'm not too much of a fan of those wooden axels because they have a lot of, for one they need to be bigger and the bigger they are, the more torque you get on the axle and the more the high friction matters. So often it's worth doing at least the axle in steel. Some piece of nail is typically sufficient because steel is harder than wood so you don't need anything fancy to make this work. But the axle one works actually pretty nicely. They are a bit more bulky than a metal hinge. But on the flip side they put the forces on a larger area of your surrounding piece so it has its benefits. Gears work reasonably well even if cut from woods. Surprisingly, wooden gears have been used for centuries even if they are typically made of hardwoods and all the gear teeth are filed to perfection. We are limited in what kind of gears we can do because we are only cutting straight down so we are limited to straight gears. If we think we would rather have metal gears we can compensate to some extent by making the gears larger. A larger gear has a larger radius which means the same torque can be transferred with less force and we can also make the gear teeth bigger which also allows transferring larger forces. So if we have the space we can afford it because the gears are much much lighter than the same gear in steel. We can also use Reck and Pinien and what I especially like is adding the gears directly to the things we want to move so we can basically buy a gear and bolt it to somewhere we can just have the thing come with an Reck or come with a gear profile. And it's a way to have relatively huge gears reasonably cheap so in one of my robots I have a 20 cm gear and it's just not an issue. You could do even bigger ones like if you have an I've seen a turntable being done that has 50 cm or 60 cm a ring gear inside there to turn that by a small motor easy doable. There's some discussion whether you want ball bearings or simple axles a lot of commercial gears actually don't have ball bearings but just have like pins in brass in my experience this works fine if you don't transfer huge powers and you basically only need gears if you want to have your output axle being able to support side loads or something like that so if you want to have a pulley which you also can laser cut with easily anhand pull you might want to use ball bearings As we are on the topic of ball bearings in time as we already told there is a great way to increase the contact area with your wooden parts you can layer use multiple layers to create a bearing seat basically have one or two layers with the outer diameter of the bearing and then have another layer with a smaller hole basically preventing the ball bearing to move inside the general rule is only use two ball bearings per axle to not over constrain them and if you really want to have precision you should pretension the bearings against each other so that to basically get out the movement and play from those bearings but that's not a laser cutting thing it's more of a ball bearing thing I thought about using wood wooden bearings basically just put a couple of balls into a recess but I think wood especially the woods we are typically using are mostly too soft for that because if you are using bearing balls they basically create point loads and wood really doesn't like that so you will get if the bearing gets hit you will get dense in your in your rays and that's just not good so maybe there's some ways this can be done but there's no obvious way of making this work that I've seen another thing that's interesting and even more experimental is using springs basically when we talk about springs we are saying cutting springy parts out of the material and using that as springs there are a couple of interesting use cases for for latches and stuff like this but it's very tricky one thing I learned is one way to make these springs have a little bit more gif is narrowing them down from their full either from their full width to a quarter or if they are in the other direction from their full height to half their height this gives about 40-60% more movement for the same stress in the same length there's a very nice document from Bayer which produces a lot of plastics that's most of this is more 3D printing stuff but that's where I found this it's a really nice document for all these bottle caps Fits and whatever another way we can use a torsion bar basically just a strip of material we can in the middle put a latch that twists basically out of plane as we discussed with torsion it's good if this has a more less scarish cross section another thing that I've used is using two buttons if you have an enclosure you can basically put a couple of long springs around one area so you can press the area down and press onto buttons for all springs it's very important to protect them from overextending so you should have a dead stop so that if you use or press the buttons they cannot break it but it will bottom out so why do I have limited success with that many materials will creep creep means the material will give under continued load so this is different from fatigue fatigue requires you to move back and forth all the time creep just means if you're putting under load it will slowly deform and turn into the new shape basically this is particularly bad with wood for some reason and but also several plastics do experience that so one way to deal with that is to basically construct around that so to anticipate that and make sure your stuff still works so this is an example that I also do have here where I have these tubs and they don't rely on them being fully extended they are basically just there to take off shear forces for the shear forces are not seen by the spring itself and also the force that's actually there doesn't really matter it doesn't rely on the strength of the spring it's just there to push these pins back out if you're putting this on top of the enclosure and of course they are also naturally limited in the way they can do they can as you can see they're basically they can't be pulled out but they can also yeah you could trap them and push them in but if you press them from the outside they are basically also limited in what a reasonable person would do to them if you have more time it's getting close flexures are a very interesting topic that I looked into that's basically a mechanism that relies on bending it's super interesting there's been a lot of fuss around this in the maker sphere I've not seen many things that are actually working well there's an open microscope project but they are using 3-Printing there's a lot of very cool and interesting stuff that can be done there but I'm not far enough to give you advice on that other than it's really cool there's a talk and workshop from Emily that you can find but there are mostly mostly indents and stuff like this so if you have a knob you can make it stop in both directions and things like that they are not that advanced as I would like stuff to be so that's something for our next talk in 10 years the last thing there will be a workshop meet up basically as an extended Q&A and a discussion thing at 16.30 which is in half an hour according to my clock in the room next door okay that's it from my side questions we have 5 minutes how is your experience with different kinds of wood or like MDF for example so we are mainly using birch and poplar and I think poplar is actually a pretty good wood the problem is it's very soft on the outside so the main problem is it's lack of hardness so it's good as a construction material but the surface will get damaged easily we have not been using harder woods than birch all at all bare lake I know there are other woods that are for example more springy like ash but we don't really have that and that's the point there's still so much stuff to do even after 10 years thank you you mentioned that a typical slicer can be used to create like a 3D model out of an arbitrary model so how does it work so does it output each individual layer or what is it I've actually not done it myself I just saw other people do it but I think it puts out all the pieces alright so just like a typical SVG yeah in my understanding I've not actually done it myself alright and I don't know what slicer they used but there was I've seen a couple of articles about that but I didn't really look too deeply into that okay interesting thank you any other questions if not there are a couple of samples in front here to look from close feel free to touch them and look at them and if you have can discuss them in the neighboring room in like 30 minutes thank you