Jump to content

geert_2

Ambassador
  • Posts

    2,084
  • Joined

  • Last visited

  • Days Won

    34

Everything posted by geert_2

  1. Could it be that the nozzle-cooling fan is not working, and the heat creeps up into and above the nozzles, weakening the filament and thus blocking further feeding? That could produce similar effects. Those fans could suck-up debris and hairs of filament, and get stuck.
  2. I think you will need to provide more specific info: which printer and model, which exact material, which settings (e.g. nozzle and bed temp, and all the rest, project file), bonding methods you tried, and especially close-up photos and maybe videos.
  3. In case of a cube, you could think of "watertight" as it being one chunk of solid metal. However, if that cube is made of six cardboard sheets glued together, with seems and gaps between the edges, then it is "not watertight". In old computer games you often see such gaps between walls, where the edges do not lign-up correctly, so these models are not watertight. Models made with SketchUp usually are not watertight either, because that program was originally made to design visual 3D-buildings for Google Earth, not for 3D-printing. For 3D-printing you need a true solids modeler.
  4. Polycarbonate is known for stress-cracking. Car headlights in PC also get dull from microcracks, so manufactureres tend to go back to plexiglass (PMMA) now. If you can find PET or PETG that can handle 100°C, it might not be the worst choice. PET should be food-safe, as it is used in all sorts of bottles and other food packages. And it prints reasonable well with decent layer adhesion and not too much warping. If the available space allows it, design the part way sturdier and fatter than the original part. Make it more "flintstone-like". And print slow, without cooling fan (if there are no overhangs), and in thin layers for best form fit, best layer-adhesion and minimal air-entrapment in the model. I never tried annealing PET, so I don't know if that is possible. PLA can be annealed, but it deforms so much, and in different ways for different brands, that it is not worth trying in my eyes. It is probably not going to fit anyway, unless you do a lot of testruns. Another solution could be: print the part as accurately as possible in PLA. Post-process for a perfect fit (but don't use it, just test-fit). Then make a silicone mould around that part. Spray the mould with silicone release-spray generously, so it is totally soaked in silicone oil, and let it dry well. And then cast some very strong filled epoxy into that silicone mould, e.g. sand-filled, metal filled or glassfiber-filled epoxy. That should give very strong parts. Chances are that your original part was glass-filled nylon or ABS too. Try to find a food-safe epoxy, or at least a not-too-smelly one, and let it cure well, preferably at somewhat elevated temperature, and clean thoroughly.
  5. Why single layer? That is going to be very fragile and might break up in steep turns or rough landings?
  6. I think you should ask that on a Bambu forum. People here are using Ultimaker and Ender printers and probably don't have any Bambus.
  7. Wow, 96-98% is pretty good. Thanks for the info. You know, in the back of my head, I still had the idea that 3D-printed metal parts were rather like sugar cubes for the coffee, sort of, but then in metal... 🙂
  8. I often find that each model needs its own settings, depending on the geometry and purpose. Some need combing, some can't have it. Some need a lot of cooling, some can't have it. Some need thick layers (e.g. for overhangs), some can't have it (e.g. for fine quality and watertight objects). Some can print fast, some can't. Some need a brim or custom skirt, some not. It all depends on the geometry, purpose, materials, fitting- and quality requirements. Even different colors of the same material might need different settings, because some flow less good than others: e.g. high-filled colors like white, yellow, lightgreen often flow less than others. If you are going to sell things, I think you should offer the STL-model (for people who do the printing themself). Or if you are doing the printing yourself, slice each model individually according to its needs, and to the customer's wishes. This for optimal quality and customer satisfaction. Otherwise they are not going to come back. Building a good customer base requires a lot of effort, it can't be done on a quicky. (That was one of the reasons back in 2014 why we bought our own 3D-printers: the parts we had printed externally, were soooo badly printed, that we could easily do better, and it was worth investing in our printers.)
  9. This is probably not the answer you want to hear, but *almost all* plastics come from biomaterial. Although some of that is a bit older, maybe a few million years... Wood and leaves fall to the ground, get covered under sand, and become coal, which can then be used to make gas and plastics. In the sea, fish and plants die and sink to the bottom, decomposing into a thick smelly goo, called fossile oil. Than is pumped up, refined, and turned into all sorts of fuels and plastics. All this carbon that was once alive on the surface of the earth, when the whole planet was green and covered in forest, has gotten stuck under the ground in the form of coal, oil and gas. Thus it is now dead carbon. So, to save the earth, we need to dig up all this dead carbon, burn it so it becomes CO2, which is then food for plants on land, and algae in the sea. Both CO2 and H2O are the most important life gasses on earth, without which there would be no life at all. Currently there is a lack of CO2, so plants are now starving from hunger. Optimally, we would need 10x more CO2 (more, not less!!!) for optimal plant growth: then plants would grow 6x to 7x faster, and all current deserts would become green again, just like in prehistory, when there was 10x more CO2 indeed. The formula: CO2 + H2O + fertilisers + sunlight ---> C-H-O-chains + O2 In words: carbondioxide + water + fertilisers + energy from the sun ---> gives plants, wood, juices + free oxygen Wood is stored solar energy. That is why it gives so much heat when burned: the solar energy is then released again. Only plants can convert solar energy into food, and convert CO2 into free oxygen. These plants are what we eat, and the oxygen is what we breathe-in. By burning the food in our body, we again release the stored solar energy in our bodies, so we get warm and we can move. We are living on solar energy, via food. In thick forests it is much cooler than on bare rock in the summer: this proves that plants do absorb the solar energy and convert it into wood. Thus more forest = cooling down. And more forest also means more food, more habitat for animals and people. The reason why forest is cooler than bare rock in hot summer, is because almost all the solar energy is captured by the trees, and converted into wood, green, and juices. So that solar energy can no longer heat up the earth, because it is no longer there, it is eaten up by the trees. That is why CO2 is a cooling gas, not a warming gas, contrary to popular belief in brainwashed people that have no clue about science, chemics and biology. They should go to school and study these sciences, and do field experiments, and open their eyes. So, to cool down the earth, we need more forest to "eat up" all the solar energy, and to convert it into wood, leaves, and juices. So we need more trees. And to get more trees, we need way more CO2, preferably 10x more. The above formula only works when there are enough ingredients to the left of the arrow. Without plenty of water, carbondioxide, fertilisers or sunlight, there is no reaction, and thus no wood, no green, and no free oxygen. Today we are at the absolute lowest level of CO2 that plants can barely live: 0.03%. Below 0.02% most plants will die soon. We are dramatically low today. So, to save the earth, to cool down climate, and to make the whole earth green again with plenty of space and food for everyone and every animal, we need to dig up and burn as much fossile fuels as possible, to pump as much CO2 into the air as possible. So we should go drive with big V8-engines that drink fuel like mad. This is pure proven science. Even drug dealers know this: that is why they add huge amounts of CO2 into their green houses, to get way more drugs. Only some politicians and some unschooled people do not want to know the facts. So I see nothing wrong in using plastics: it helps making the earth greener, producing more trees, more food, and more habitat for animals and people, and it helps cooling down the earth. So don't feel bad about it. As I said, probably not the answer you were looking for. But it is the proven scientific thruth.
  10. I think you need to keep your nozzle cleaner. I always wipe it immediately after a print completes. If there is stuck residu, I gently (always very gently!!!) scrape that residu off with a very soft brass screw thread, whereby the treads act like a soft file. Never use steel or hardened things, that could damage the nozzle. Sometimes I oil the nozzles when hot, with PTFE oil. This reduces the amount of material sticking and accumulating onto it, but it does not totally eliminate the effect. What can also help against edges or corners lifting, is using a brim. Or use a skirt that is very close to the model, e.g. with 0.1 to 0.3mm separation: that separates easier than a brim, but still helps in keeping corners down. Or for stubborn cases, design a custom brim into your model in CAD, in such a way that it is very stable, but still easy to remove. This may require quite a bit of trial and error, on small test pieces, just to find the optimal concept. Apart from this, I have no experience with your material. Maybe try different glues or adhesion-sheets (like wiping dissolved wood glue onto the bed, tape,...). For my PLA and PET prints, I just *never* use glue..., so I have no idea about your material.
  11. If you print them for real use, and not for demo-purposes: aren't they too brittle to screw into wood? I even occasionally snapped standard bolts and screws, before I began to use copper grease on metals, or a drop of oil on woods for lubrication and thread cooling.
  12. I also print PET (or PETG, I don't know) at the lower edge of its temp range, sometimes below its recommended range, and at far lower speeds than PLA. When molten, PLA becomes sort of yoghurt and flows easily, but PET rather stays like soft, rubbery chewing gum, it does not like to flow well. PLA bridges well, but PET tends to snap and fold back onto itself, so instead of a nice bridge I often get a sort of "grapes" accumulating on the edges. And as said before, it tends to accumulate on the nozzle, causing a thick blob that slowly sags onto the print every so many minutes. The blob gets brown due to decomposing, if the temp is too high. I use no cooling fan, and have no problems with warping. I find PET is best for things that need a little bit of flexibility, like snap-fit mechanisms. But it is not stronger than PLA, only more flexible, and has a bit less creep under load. And it can go up to 70°C where PLA can only go to 50°C, and under load even not that. So PET is suitable for use in a car, PLA not (don't ask how I know). When printing without cooling fan, layer adhesion is good, I have no separation. Clarity in transparent PET improves a lot when printing slow and in thin layers, indicating that there is far less air entrapped between the extruded sausages. All bad effects are minimised when printing cooler, slower, and in thinner layers. A few examples: Ruler is in cm an mm: PET can be chemically smoothed with dichloromethane easier than PLA, and it does not seem to dry out in het months after treatment, contrary to PLA that tends to develop microcracks. Both parts are identical, but one is smoothed to remove layer lines: much more hygienic for use in hospitals, easier to clean. Transparent PET printed at different speeds and layer thicknesses: blocks are 20x10x10mm. Printed at 0.4, 0.3, 0.2, 0.1 and 0.06mm layers, top row at 50mm/s, bottom row at 10mm/s if I remember well. The yellowish discoloration is from sitting long in the nozzle at very slow speeds: it begins to decompose. Temp mostly is 215°C, except the thickest layers at 225°C, and the thinnest at 200 or 205°C. The watermark with my name is sitting halfway inside the block, it's hollow text, and only readable at low speeds and thin layers.
  13. Youtube is a good place to find tutorials, especially for very specific objects. Search for "cad how to design a [object] in [3D-program]", and similar. For the free software DesignSpark Mechanical (from RS-components), you can also use the basic manuals of Spaceclaim, since DSM is a limited version of Spaceclaim.
  14. Indeed, as gr5 says, preferably post pictures instead. Most people do not trust zip-files because they could contain virusses or hacking software. That could be done on-purpose by hackers, but also happen by accident without the honest poster knowing. So a lot of people are not going to risk opening a zip-file. I am not saying your files might contain anything irregular, probalby not, but I am not willing to take the risk.
  15. I have seen this occurring when the extruded "sausages" do not bond perfectly to the glass, or when there is a bit of overextrusion, or the nozzle too close to the glass (so the melt spreads out too far). The sausage is layed down but its edge curls up just a tiny bit. On the next infill pass in the opposite direction, these curled-up edges are remelted and pushed down again onto the bed. If only part of it curls up, and part not, then you could get this effect too. Because the originally layed down sausage and the remelted and pushed down again edge spread in a different way, and overlap in a different way. This happens if bonding to the glass is not equal over the surface. But you have to keep watching very closely while printing, to see it happen. You can see the effect to a lesser degree in these two prints: a tiny bit in the first model... ...and a bit more in this model. Note: the vagueness and indents are because of tests with solvents that weakened and deformed the model, making the circular indents on the other side shine through. I am not sure it is the same effect in the topic starter's photos, but it could be?
  16. Out of curiosity: why print upright instead of laying down? Resolution may be worse when laying down (i.e. about 0.5mm with a 0.4mm nozzle), but I would expect the "flatness" of the whole image to be much better? Thus there should be no layer lines at all, since it is just a thin plate? So that should result in a cleaner image in see-through, I would think? Or am I missing something? I tried a few small lithophane portraits that way (laying down) on my UM2 in a translucent PLA (colorFabb's "natural" spaghetti coloured PLA/PHA), and the result was quite nice: it was weird when looking directly upon it because the eyes are like sharp spikes protruding, but it looked good in see-through. But it takes some testing per material to find the best thickness, because that depends a lot on translucency. Also, I found that I needed to print slow, cool and in very thin layers, not to entrap air in-between the extruded sausages, to avoid the diagonal infill pattern lines.
  17. That depends a lot on the model, printing speed, material and temperature. It may also depend on your slicer. Smaller models can get away with smaller clearances. But in bigger models, this makes it harder to remove the support. Also, the easier you can get in with tools, the narrower the gap can be. So it boils down to trial and error: make a testmodel with various clearances that mimicks that part of your real design (like the one shown above with the numbers 0.2, 0.3, 0.4 and 0.5mm), and print that with your desired materials, speeds and temps. But generally, clearance is indeed between 0.2mm and 0.5mm, with 0.3 and 0.4 being most common in my designs. And it also depends on the functionality: if accuracy does not matter too much, take bigger clearances. If accuracy is important, for example in rulers that have to slide into each other, then tighter gaps are better, but that requires more effort to design and print the model and to remove the supports. Thus provide enough features in your design to make support-removal possible, so you can get in there with a knife, scalpel (=surgical knife), hook, pliers, or whatever else tool. A massive block of support in a tight area is almost impossible to remove. The better you can get in with tools, the tighter you can make the tolerances. For example to wiggle-out supports with pliers, extend the support to outside of the model to get a good grip with the pliers. It is worth thinking this over very well, and to make several simple concept tests, before doing it in the real model. Especially if the real model is big and complex and takes a lot of time to print, or if you have to print a lot of them. That extra testing time is very well spent. Keep watching while these test models are printing, and look what happens. That also gives you a lot of insight, especially for the free-hanging supports, because of their tendencies to sag and to curl up at the same time. So you can catch and correct any outpoints immediately at the beginning of the design phase.
  18. To me this looks like intermittent underextrusion, most likely? So, check the whole path from spool to nozzle: no partial clogs, no worn-out or dirty feeder wheels or other parts? No stuck filament on the spool (can happen if you have a spool where filament loops under each other), no kinks or bends in the filament which cause huge extra friction (can happen in a poorly wound spool, or a rewound one), etc...?
  19. To me it looks like infill is shining through, because in the first pic it seems to repeat very faintly in the rest of the print too? Or is that just my brain that repeats the pattern, after staring for too long at it? If it had been on the bottom layers, I would guess a greasy dirty build plate, but that can't be on top layers obviously. If a printer would have mechanical issues, like linear bearings with defects plus rods with defects, I wonder if that could echo into the print like this? Seems far fetched to me, but maybe not impossible? I often add watermarks to my prints, thus hollow text inside a print, but that is clearly visible in the slicer, if you set it to transparent mode, or layer-by-layer view. And it is definitely not random. Any model-defects like double faces, or minor indents, should also show up in the slicer preview.
  20. No, I haven't figured out what it is. It is totally random, sometimes it immediately happens when starting to rotate the knob, sometimes after a lot of rotations (e.g. when manually dialing-up the temp from 0 to 200°C), sometimes not at all. It can happen in all menus where the dial has to be rotated. I don't see any logic in it. But indeed, when the printer is printing when the display locks up, then it keeps printing. So it is not the main system that locks up. It is the display module or interface to it only, for whatever software- or hardware reason. The total randomness when rotating, makes me think it is a hardware thing, or interference thing, but not sure. It happens only on one of my UM2, the other one is fine. Both are from the first UM2-series, non-plus, non-connect or whatever the newest versions are named. But there is a difference in firmware between both, and probably in hardware too: 1) The first one is okay (=firmware/hardware where the little nozzle-fan always runs). This is firmware version:_14.12.1 of date: 15 dec 2014, 15:03:28, printer serial nr.: UM2A91 -MES133705 2) The second one has the occasional display-lockup (=firmware/hardware where the little fan only runs when the nozzle is over 40°C). This is firmware version:_15.02.1 of date: 19 feb 2015, 09:59:45, printer serial nr.: UM2B56 MES133716 I don't know if I could transplant the older firmware from the good one to the newer not-so-good one, because of the differences? Or where I could find each firmware version from that time, to experiment? But I have not really looked into it either, because it still prints, and I don't want to repair it until it is broken... :-) There is also a minor difference in printing quality. Maybe because the bed temp seems to be regulated differently, or the board layout is differently? In the second one, heating the bed seems to influence layer-thickness or nozzle temp or melt-viscosity a little bit, not dramatic but just barely visible, causing faint horizontal lines similar to when you dial the temp manually on the fly. Also not sure if this is hardware or firmware. My older printer does not have this. Is there somewhere a list of which printer versions or serial numbers have which hardware main boards? And which are the differences? @gr5: I don't know what geek-mode is? Was that already on those old UM2 printers from late 2014 / early 2015? And if so, where could I find it? Or is that something on the newest printers only? I am only operating them with an SD-card, never tried via USB. I also thought of adding ferrite rings around all cables, to reduce the chances of interference, but haven't done it yet. Could also be a weak power-supply, on the edge, causing glitches? As said, I am just guessing... :-)
  21. Forgot to say earlier: the height of the text (= how much it is raised or recessed from the surface) is also important for legibility and quality: too thin and it will not be legible, too thick and it will get more deformed. You might also want to play with that. Usually, I make it raised 0.2 to 0.3mm, or recessed 0.5mm. Hollow watermarks look best around 1mm thick, sitting 0.5mm below the surface. When buried too deep, watermarks will be blurred and unreadable. Watermarks can be (1) hollow text, or (2) solid text surrounded by a hollow rectangle: both give a very different appearance, with the solid text in a hollow rectangle usually looking better. Both versions are shown in my testplates. Small recessed text is most problematic: the nozzle can not get into thin openings of characters like N, M, W,..., so these voids tend to get closed. See my test-plates with different heights of raised and recessed text, and different watermarks (via URL above). You can simply print these STL-testplates, and see how it comes out. Each STL-file is accompanied by a JPG showing what it should look like. Further, the reason for printing slow is that this gives less ringing and less overshoots, less thickening of corners due to braking, and less variations in nozzle-pressure, and thus less variations in extrusion-rate, and thus smoother letters with less defects. The reason for printing cool is that this gives less stringing, and a more accurate shape, as long as you have no underextrusion. Thin layers also give a more accurate shape, where in thicker layers the text looks rather like bent sausages. The bright yellow plate has solid watermark text surrounded by a hollow rectangle, caps height 3.5mm.
  22. The most important thing in printing small text is to print it very slow, in very thin layers, and as cool as possible. But with a standard 0.4mm nozzle, it will always look like from the Flintstones, simply because the nozzle can not make sharper corners than 0.4mm, and it can not get into narrow openings in letters like N, M, V, W. Also, characters like 8, B, D, O, Q, tend to get rounded off and look rather similar. But it is legible, and good enough for copyright-text, on/off switches, rulers, and similar simple indications. I also like to use small text for watermarks in transparent and translucent materials. Watermarks can have a bit rougher shape, that is to be expected. During the last years, someone made a real font from my vectordrawings shown above, so you can find links to that on my personal page (scroll down there): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ Below a couple more pictures (also see the ones above). Unless otherwiste noted, all caps height is 3.5mm, character-width is 2...2.5mm, and leg-width is 0.5mm, all printed with a 0.4mm nozzle. So this is what you could expect from "normal printing" without any special equipment. Raised text: Watermark sitting 5mm deep halfway in a 10mm high transparent block (left: as printed; right: after some polishing): More watermarks, design files: Usually I place watermarks ca. 0.5mm behind/below the surface. That keeps them legible. Watermark, also note the 0.4mm wide diagonal layer-lines: This is ultra-small text, 2.5mm high and 1.5mm wide, instead of my usual 3.5mm. This is the limit of my printers. Very much "Flintstonian": I make the watermark by first designing the text separately, thus outside of the model. I save that as a separate model for later re-use. Then I move the text inside of the model. Upon exporting to STL, my 3D-editor DesignSpark Mechanical automatically subtracts the text from the model and produces correct STL-files. So I don't have to do the subtracting myself for a fully submerged watermark. That makes it a lot easier to later edit the text. In case of raised or recessed text on the surface, obviously I need to do all merging myself, since we can not have partially overlapping models. The merging should be done only as the last step, after all other editing is completed. This below is fully submerged text, unmerged: More watermarks: the top model is smoothed after printing, so the layer-lines are less visible and the watermark shines through better. The bottom model is as-printed, with the layer-lines reflecting light and hiding the watermark a bit. Raised text: Separate file with my usual copyright-text, so I can easily import that into any design, move it to the right place, and (if raised or recessed text) merge or subtract from the design. Font shape is very crude with only straight lines, which makes rendering and slicing faster, conserves memory, and keeps files small. Making it more detailed or rounding corners makes no sense, since the printer can't print those details anyway.
  23. I have the feeling that you printed this a bit too hot and too fast. Also, white often does not print too well, I am not sure why but I guess because it has fillers to make it opaque white. Further, older filament might be moist, or it might have become too brittle or too hard to feed well. If the spool is nearly empty, a hard and tightly wound filament will cause a lot of "unwinding resistance" (=it acts like a spring and wants to wind-up again) and cause a lot of friction in the bowden-tube and nozzle, causing under-extrusion. This is a known issue with the UM2. If so, you could unwind a bit manually, bend it in the opposite direction around a skater wheel, release again: this will straighten it somewhat, causing far less friction and resistance. For best quality, print slow, in thin layers, and cool (=at the bottom of the temp-range of the material), with new filament, at the beginning of the spool (or if at the tight-wound end, after straightening the filament). For example, printing 0.06mm layers, 25mm/s, 190°C for PLA, will give good quality, but will cost an endless amount of time, 12x longer than printing at 0.3mm and 50mm/s... Most of the time I print PLA at 210°C, 35-50mm/s, and 0.1-0.3mm layers, with decent quality, good enough for the purposes. If you are relatively new to 3D-printing, or on an unknown printer, make a small test model, and print that with different settings. Or change the settings on the fly while printing (temp and speed), to see the effect. Pictures: Testprints at various speeds and layer-heights: from left to right: 0.4, 0.3, 0.2, 0.1, and 0.06mm layers; top row at 50mm/s, bottom-row at 10mm/s. Material is PET, transparent to see what happens inside. The watermark is halfway inside the model. Test-block dimensions are in mm: 20 x 10 x 10, hollow watermark text is 3.5mm caps height and 0.5mm leg width. High-quality print with 0.06mm layers and ca 0.35mm/s (if I remember well). Ruler is in mm and cm. Standard quality print at 50mm/s. Unwinding and straightening a bit of old, stiff and too tightly wound filament, by bending it around a skater wheel in the opposite direction, then releasing it again. After that, its bending radius is about 30cm, the same as that of the bowden tube, thus causing almost no more friction and no more unwinding resistance in the feeding system.
  24. PLA under load, in my experience, will cause several problems: - creep deformation, as mentioned above: the material slowly deforms under the constand load, and clamps and screws will come loose, - micro-cracks will form in areas that are under constant tension, - from ca. 45-50°C onwards, the part will deform due to heat, so it is not suitable for use in a car or machine for example, I have seen all these effects in my own models. I don't know about sand-filled or metal-filled PLA, no experience with that. If it needs to be really strong, what I would do for myself is: - print a mould in PLA, and pour a strong sand- or metal-filled composite in it. - or print a master model in PLA, make a silicone mould of it, and then pour the sand- or metal-filled composite in that silicone mould: that will make releasing it a lot easier. This method with silicone moulds is also used to make art statues in cement or filled-epoxies. Thoroughly soak the mould with silicone oil prior to casting, that will make it last longer. Remember that silicone moulds are water-tight because they repell water, but they are not solvent-tight, and not oil-tight: these will penetrate the silicone. Composites will also penetrate the silicone, cure in there, and thus damage the mould over time. That is why you need to saturate the mould with oil first, so no other products can penetrate it.
  25. I don't try to print threads in PLA anymore: if they show up at all, their shape is not conform specs, and they are too weak anyway. When fastening a bolt, you have to force it in (because of the out-of-spec dimensions), and this creates enough friction heat to melt the PLA and destroy the whole thread... When possible, I now provide room for a hex nut, and I use standard M4 nylon nuts and bolts. If the opening for the nut has a "metrinch" shape (=hex with curved sides), you can make it so that the nut sits really firm into the cage. Below a few examples: Testmodel to see which dimensions suit the nut best (size in mm, for a nylon M4 nut): "Metrinch" cage for an M4 nylon nut. The conic top is to make printing easier without the top sagging into the cage. The flange at the bottom is to make it easier to insert the nut. Make the shape tight, so it sits firmly. Bottom view of the cage: Same concept, but this time with an extra retention ring to hold the nut firmly in place. Design the nut first (pink here), and save that as a separate model for later re-use. Drill a hole in the model where the cage has to come (yellowgreen), align the nut to that hole, and then subtract the nut from the model. So now you have a model with cage (cyan): Another way of using standard nuts and bolts. This nut slides in from the side, and can not fall out as long as the bolt is in place. If the hole is tight enough, or with retention features, it will sit very firmly anyway: Another model with M4-metrinch cages for standard M4 nuts. Due to the tight metrinch-shape, I have to hammer-in the nuts: they won't fall out, but can be pushed out if required:
×
×
  • Create New...