Jump to content

geert_2

Ambassador
  • Posts

    2,084
  • Joined

  • Last visited

  • Days Won

    34

Everything posted by geert_2

  1. The first photo looks like poor adhesion to the bed,. Maybe due to: - Nozzle too high? If so, adjust bed-leveling so that gap is thinner. - Dirty, soapy, oily bed? If so, clean with isopropyl alcohol, then with pure tap water. - Maybe other reasons? So I would first look into that direction. The second photo, I can't really say.
  2. If it was my core, I would probably: - first follow all guidelines on hot and cold pulls that I could find, - google for more info and videos, - then I would consider using a long and thin brass wood screw, to try to extract bits and pieces of junk. Or modify a brass screw, so it looks more like a drill bit. But it has to be a very soft brass, so it does not damage the inside walls of the nozzle. Maybe you need to find out how to soften brass first (maybe in a flame, and then slowly cool, I don't know?). Use it manually to get a good feel and go slow, definitely not an electric drill. - then I would consider burning it out by prolongued heating of the nozzle, - and only then I would consider disassembling it I think... But this is my personal view, not an official recommendation, as I am not related to the Ultimaker company.
  3. I have no experience with the mould-option. But why not make the mould yourself in CAD? Then you have all the freedom you want, and you can insert any desired features. - Draw a box around your model. - Subtract the model from the box, so you have a hollow (mould) shape of the model. - Provide pouring openings and funnels. - Provide venting openings for air to escape. - Provide flanges or screw-openings to assemble and clamp both mould halves later on. - Design a thin plate and provide alignment features on the edges it: blobs on one side, pits on the other. - Use that thin plate to cut the mould in two halves. So both halves of the mould will align correctly later on, thanks to the blobs on one side, and pits on the other. - Print both mould halves on their back. That is the advantage of using a box to start from: then you end up with two halves with a flat backside, easy to print. Print with wall-thickness and support-density as required for your castings. Concrete and gypsum will need stronger parts than foams, obviously. - Spray the mould with suitable mould-release spray for your materials. Then spray again. - Clamp both halves together, and ready to go. Then you have a professional 3D-printed mould. Very similar to hand-made moulds. All these features are very difficult to generate automatically; you can design them better manually. Watch out for exothermic casting materials (even gypsum): they might melt the mould. Use slow-curing or non-exothermic materials.
  4. Maybe use a flexible PU coating? PU tends to glue well to most materials, and is quite resistant.
  5. Or replace an old cartridge just before it is empty, and test using the remains, then you don't waste less. Coating with non-stick teflon usually requires hot temperatures to bake it on. Otherwise you are back to oil-based teflon sprays... Depending on whether the 3D-prints are end-products, or intermediate, or tools like clamps, another option might be to make a mould from 3D-prints, and then pour a suitable material in it, for example silicone like used for mould-making and film-props. This is non-stick silicone that comes in various hardnesses, from skin-soft to tire-rubberlike. Or 3D-print a PLA-original, make a mould around it, remove the PLA-original, and pour the desired material in the mould. Or if you only need very few items, have them machined from blocks of teflon, PP, PE,... Prototyping companies like Protolabs might be able to do this.
  6. Just wondering: while printing, can you see it stuttering? Or if you put your finger on the head, can you feel more vibrations in the head than in the other printer? And if you put your finger on the feeder, or the filament just before the feeder: do you feel (more) stuttering? This might give a clue whether it is mechanical in the head drive train, in the feeder traject, or electrical due to temperature changes, or power supply changes, or interference or something?
  7. My experience is similar: print as slow and as cool as you can, and in thin layers. Also try the opposite, print fast, hot, and in thick layers, and see how that works. By exagerrating (here in the wrong direction) you get a better understanding of the effects of these settings. Stay with the printer and watch what happens. You can also adjust settings (speed, temp, flow rate,...) on the fly up and down, to see how they affect the print. Keep watching it: this gives a lot of understanding. Obviously do this on small test pieces that don't take too much time and material to print.
  8. There is a post of me with more tips for single nozzle - single material supports, but I can't immediately find it. Anyway, here a few of my standard pics with tips: Tiny custom support of 10mm wide (dark blue) for the overhang. Ribs on top (0.5mm) allow tighter gaps, but still prevent the support from sticking too hard. Supports as thin layers, not connected. When printing they will sag a bit, but can be peeled off one by one. Usefullness depends on the model. Free hanging support concept for overhangs. The inverted staircase reduces the tendency of triangles to curl up. The tiny distancing pins (1mm gap, pins are ca. 0.5mm wide, 0.2mm high) allow easy cutting off, without much damage to the model. The 1mm gap is to be able to insert a knife or scalpel. Idem. This concept of free hanging supports reduces the amount of support needed, and they do not destroy underlying parts of the design (see my name), contrary to vertical support-columns. The orange and pink parts are supports: they extend from the model, so I can grab them with pliers and wiggle them loose. And they have extra brims to improve bonding, as they are very small, too small to get in there with a knife or scalpel to remove. For reference: the text caps-height is only 3.5mm. A few more concepts: separate chunks of support, so you can wiggle them loose individually. Extended supports, and supports with holes, so you can insert a tool to pull them out, or grab them easily with pliers. Slightly bigger supports than the model (bottom right) to improve the outer edge of the bridge. Make a test pattern with different methods and gaps to see what works for you: it depends on materials, models, size, and printing speed and temp, so you have to try.
  9. So you mean that you get underextrusion on long lines of infill, and overextrusion on very short ones? I believe the overextrusion is caused by the nozzle having to slow down in order to take the 180° corner at each end of a line. But the extrusion rate keeps going, because pressure in the nozzle does not immediately drop to zero, so it momentarily overextrudes at corners. Those very short lines consist of nothing but corners. Printing a lot slower, and a bit cooler, should improve this but won't eliminate it. There may be other settings you could adjust, but I don't know them. I notice this effect too when it is printing the short infill-paths between the holes. Although it is hard to see here in the photo. Here, this effect is slightly visible on the right edge of the right orange model. (The left- and center models are post-treated, the left with acetone-smoothing, the center with heat-treatment, so they don't count.)
  10. Usually nylon, polyethylene, and polypropylene are very hard to glue. Also PET can be somewhat hard to glue. Maybe you have a chance with these? But I haven't printed with the first three, only with PLA and PET, so I don't know if and how well they print on an UM2. Before buying filament, maybe you could test what the UV-ink does on cheap parts in these materials? E.g. PE or PP boxes for the fridge, nylon screws, PET bottles,...? On a 3d-printed part it will probably stick a bit more, due to the layer lines that give more grip. Another option could be to spray the PLA with mould-release spray, silicone oil, or PTFE oil, if your application allows it (outside of the printer!!!)? These can be found in car-accessories shops, brico shops, or art shops.
  11. If it is lost anyway, I would also try to disassemble it, even if only to see how it is made.
  12. I think that "ring" on your last cold pull pic should probably not be there? (See the very left border of the pic.) Long ago I have seen a short video of a coupler that did not fit well, and molten filament came in-between nozzle and coupler, pushing the coupler up like a hydraulic cylinder. I don't remember the cause, but this could happen if there is burned residue sitting between nozzle and coupler, so there is a gap. You might want to check that? But I also think you should double-check if the third fan keeps running well while printing. Worn-out or dirty fans might start well, and then slow down or stop, or vice-versa. Not saying this is the cause, but I would double-check it. For nozzle-cleaning, I also think you should wait *much* longer for the nozzle to cool down, so that the inner core of the filament totally solidifies and comes back to room temp. Then wiggle and try to rotate *gently*, before pulling. This helps dislodging dirt from the nozzle. Then I reheat the nozzle to about 75°C (for PLA), while *gently* pulling and rotating the filament, until it comes out. I never use brute force, only very gentle wiggling, rotating and pulling. And longer waiting cycles. See my very old manual on atomic pulls here (I should update it...): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ Typically my cold pulls look like these below. The bottom orange one is best. For the top one, I did not wait long enough, the core wasn't totally solid yet. If there is burned residu on the walls of the nozzle that I can't get out with pulls, then I *very gently* scrape the inner core with an M3 threaded rod, with rounded end to not damage the core. Never use steel: this is too hard for brass nozzles. Very gently scraping; this is on an UM2. Not sure if this is a good idea on other printer models though... Also note that this is not an official method, just my own thing; I am not related to the Ultimaker-company and not an official spokesperson. Other puls: the white one shows a totally worn-out teflon coupler (UM2), indicated by the thick blob or "ring". This does cause underextrusion too, and needs replacement. If there would have been a gap between coupler and nozzle, the "ring" would have been where the dark line now is on the white one. On an UM2, then the whole nozzle would need to be disassembled to clean that (I don't know about other models).
  13. To me it looks like you used a bit too much glue stick. Only use a little bit, and then spread and wipe it smooth with a wet tissue. For printing PLA, I do not use any glue at all, nor any blue tape. After thoroughly cleaning the glass, and then cleaning again with warm tap water only (no soap, no detergents), I wipe the glass with a tissue moistened with salt water. I gently keep wiping until it dries in a thin mist of salt stuck to the glass, almost invisible. My "salt method". See here for an old manual: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ For me, this salt method greatly improves bonding compared to printing on bare glass, when the glass is hot (60°C). When cold, there is no bonding at all, nothing, so the models come off by themself. No need to take the glass out of the printer. Then, for the next print, I wipe the glass again. This works very well for my low, flat, long models. But it is not recommended for thin, high models like lantern poles: they may come off. Other people use the glue stick (and wipe with a wet tissue afterwards), dilluted wood glue, hairspray, 3D-LAC, or similar. Try on a small test part what works best for you. And in the beginning, stay with the printer, and keep watching what happens. Until you have a method that works flawlessly for your prints and materials. A test-model to try-out bonding to the glass. These inverted prisms are a very hard test. If a bonding method works for this, it is likely to work for most other models too. Note the salt on the glass: this is a bit too much, less would be better. My salt method works well for such long flat models, but not so good for high models like lantern poles. Bottom of a test prism. Some corners lifted, but it still could be completed.
  14. Thanks for the good idea. I have one glass upside-down. But this may come in handy one time if another chunk would be torn-out.
  15. So I made a nice pic to illustrate this concept of lubricating the bearings of old fans. This is a 386-computer fan, not an Ultimaker fan, but the concept is the same. I have done several of these in the old days. Feel with your fingernail where the edge of the bearing is. Mark this with a pen if necessary (see red circle here). Then inject a little drop near the top (thus near what would be the top after mounting it again). A little drop, not a whole syringe: you don't want it to spill oil on the glass and destroy bonding. Somewhat thicker bearing oil works best in my experience. Too thin oil leaks away, does not fill the worn-out gaps, and does not stop the shuddering of the fan. Don't do this on new fans: they come lubricated optimally. Only on old worn-out ones, once they start running rough and making "rheu-rheu-rheu" sounds. And only after checking that the problem is not caused by ingested dirt, hairs, strings,..., or anything else, but really the worn-out bearing. Then buy replacement fans, as this method may work for several months, but not forever. Once you find that you need to re-oil every week, then it is *really really* time to replace it. :-) For safety: use a blunt needle, or cut the sharp tip off the injection needle prior to use. You don't want to accidentally inject this stuff into your body. Cutting the needle dan be done with a Dremel tool and cutter disk (use safety glasses!). Obviously, don't use this method on life-saving equipment, critical systems, aviation systems; only on sturdy systems that won't burn down when overheating a bit.
  16. Just out of curiosity: when handling resins, do you do that in daylight, or in a dim-lit room, or an almost dark room? To prevent them from self-curing? How sensitive are these resins, let's say if you leave them out in daylight or in the sun? How much time do you have before they self-cure? I am asking this because I have worked with dental light-cured resin pastes: the white filling materials for teeth. And you should definitely not use them for longer than half a minute in bright light.
  17. My experience: try to avoid self-tapping screws, or thread-tapping, in PLA. Tapping is likely to melt the plastic and destroy the thread, even when tapping/screwing very slowly and carefully. At least in the sizes M3 and M4 that I use most. The heat generated by the friction and deformation from tapping immediately goes above 50°C. Also, PLA is sensitive to creep-deformation under load, so it comes lose soon. It might work with other materials like PET, but I haven't tried that yet. This is much less of a problem with inserted real nuts/inserts. Although I haven't tried it myself yet, this is why I believe that heating the nut/insert as Oliveros said, and then sinking it in, is a better solution indeed. After cooling and solidifying there will be less rest-load, and less tendency to crack and deform. By the way, I prefer nylon nuts and screws for plastics, when possible, as in the pic above. These don't fall out easily, and are a bit flexible. This makes fastening PLA parts more reliable and softer, not so On/Off as with metal screws, and it withstands creep-deformation better for low force applications like the clamps above.
  18. I have also been wondering about that. I have seen the toy a couple of times, yes. But I think with a bit of searching in a hobby-shop or brico-shop, people should be able to find replacement wheels, like the standard wheels for shopping carts, or chairs,... Anyway, 3D-printing them will be an excellent test of layer-bonding. If kids can't destroy it, then nothing will. :-)
  19. Not an official answer, but based on common sense, I think the problems are not going to be the plastics. Kids wear plastic rings and armbands all the time, and most clothing is from plastic these days. But it will rather be the tiny openings in which bacteria and dirt can get a grip. So you would want to make your prints as smooth as possible, with as little tiny openings and layer lines as possible. PLA is bio-degradable, so theoretically bacteria could eat it. But I have printed PLA sifts for the sink in my laboratory, and after years of use in the wet and dirt, they don't seem to be affected too much. So, unless a person is very sensitive and has contact allergies, I wouldn't expect problems from the plastic. But maybe from dirt/bacteria in the layer-lines.
  20. Another thing I forgot: smooth the mould prior to casting the silicone. Every detail will be replicated, so layer lines will be visible in the silicone too. This will make removal of the part from the mould more difficult, and it will make it more difficult to clean. So, sand, polish, chemically smooth ("acetoning"), or paint the mould. This is well spent time.
  21. Yes, if you have the tools, I can imagine that heatsinking-in real knurled inserts is the best. Probably the strongest also, less likely to go rotating. For completeness, if the design allows it, and if forces are low, inserting the nut via side-openings might also be an option. See the pic below. I use this for light clamps. Advantage is that this is very easy.
  22. This is a good This is a good idea, worth remembering. This concept could be usefull for lots of other models too.
  23. One extra note: if you use a thin, slow-curing silicone for casting, be sure that the lower half of the mould is absolutely water-tight. Silicone slowly creeps into the tiniest openings, even only microns wide, and would leak away. With thick, fast curing silicones, this is less of a problem, as they are already cured before they have time to leak away. If there are non-watertight seams, you can close them with plasticine or wax. Another option is to make the whole mould out of plasticine. Be sure to use non-sulphur plasticine, as sulphur inhibits curing of silicone. I have used both 3D-printed moulds, and plasticine moulds, and combinations. Obviously, you need non-stick silicones for mould-making, not the sticky sanitary or construction silicones.
  24. The little rear fan (nozzle-cooling) should always be on when printing (when the nozzle is above 40°C). If not, check if there are no hairs and strings of filament or dust stuck in it. Check the connections and wires. If the bearings would be worn-out, or the wires broken, replace.
  25. If you say the fan is "making noises", it could be: - There are filament hairs and strings stuck into it, hindering the rotation. - Or the fan bearings are worn-out. This was very common on similar CPU-fans in older computers (286-386-etc.). If you don't have a replacement yet, you can extend the life of the fan a little bit by lubricating the bearings: using a needle, cut through the silver label covering the bearing, and then inject a drop of bearing oil into it. This is obviously not a permanent solution, but I did use this quite often on computer fans in the old days, and it can extend fan life for a couple of months. Enough to keep working and meanwhile find a replacement fan (and time to install it).
×
×
  • Create New...