Jump to content

geert_2

Ambassador
  • Content Count

    1,386
  • Joined

  • Last visited

  • Days Won

    17

Everything posted by geert_2

  1. Another option might be to make a mould, and cast it with a high-temp material, glass- or fiber-filled, or non-filled, whatever you desire? There do exist high-temp polyurethane and composite materials. You could 3D-print the original model, clean it up until perfect, and then make a silicone mould of it? There are lots of good video-tutorials on Youtube. Search for: moulding and casting Also, depending on the vibrations and load-spikes, I am not sure a snap-fit mechanism would hold. It surely wouldn't if mounted on a diesel engine block at a rough coldstart. In such a case you might want to provide additional clamps: metal spring clamps, multiple nylon cable binders (but they tend to fail under heavy load), self-locking screws, or similar common automotive clamps? If you could avoid the snap-fit, you have no problems with flexibility requirements for it.
  2. Yes, this seems like a very good idea, I hadn't thought of that: less friction, and far less complexity, especially when changing filament. You just need to make sure that the external tubing is long and flexible enough to be pulled up to the feeder automatically when the filament moves.
  3. I think you can clamp the bowden tube with pneumatic clamps/fittings. I guess the bowden tube is a standard size, but maybe imperial? Your tubing might have a different outer diameter. And then you would need to design some sort of housing for that clamp, and mount that housing next to the feeder. See the pneumatic fittings in the image below (the ones with a thread).
  4. Watermarks do reduce strength of the model, obviously, since there is no material there, and the gaps might concentrate stresses. Apart from that, I never had issues with internal watermarks. Except if the hollows are too big, and their "roof" doesn't close very well: then you see some spaghetti or grape-like blobs sagging inside. And if they are too deep into the model, they become invisible or blurred, due to the layer lines acting as a blur filter. But this is just cosmetic. In my keychains above, it hasn't caused any mechanical or software-trouble. Originally I made watermarks by subtracting the watermark text from the main model. But this proved unnecessary. It is sufficient to just move the text inside the model, without any further operations. Then, upon export to STL it is automatically subtracted correctly, since STL are just triangles without any meaning. As far as I understood, "watertight" has nothing to do with hollow design features like watermarks, but with the edges of the model touching each other correctly. If you fold a paper model incorrectly, it has gaps in the seams, so it is not watertight. The same in STL: edges of triangles not touching each other. I made these keychains in DesignSpark Mechanical, and exported to STL. Don't use SketchUp: this causes problems with watertightness indeed: its vectors don't match, and they don't close gaps in the seams. Concerning the reinforcement of holes, I think this is possible by defining a different infill percentage around the holes. But I don't know the details, so I will leave that to others.
  5. My experience is that usually I have to drill out holes anyway, because of occasional little blobs, strings and hairs. I also tend to make them 0.5mm wider in CAD than necessary: thus 4.5mm for a 4mm hole to slide an M4 bolt through. And then I *manually* drill it out (see photo). This only takes a few seconds. No electric drilling, especially not in PLA, because it melts immediately.
  6. I have seen this problem too in long prints: while printing, the model sticks to the glass very well: no lifting/warping at all. But after removal, it starts to deform slightly. And indeed, for PLA usually the center "lifts" half a mm, if you put it on a flat table. So this must be due to residual cooling stresses in the material, which are not enough to pull the part off the glass, but enough to slightly bend it. However, I don't have a good solution: warming it up again to release the stresses, tends to make things worse, far worse. I have had models warp upwards on the edges. Or others warping upwards at first, but then after some time reverting and severely warping in the other direction, very weird. And they shrink in length, but expand in width and height with several percent. Maybe, if you could clamp the model between heavy metal plates, and gently warm it up to around its glass transition temp, and very slowly cool down, you could remove some of these stresses? Or let it sit for a night on the warm glass bed of the printer, before removing it (and cover the bed with a box, so temp is uniform)? Edit: the photo shows some of the warped test models. In this case they were subjected to way more elevated temperatures after printing, to try the effects of annealing. Thus way above glass transition temp. But the bending upwards in the center shown in the second model, is also very lightly present in 20cm long models, immediately after removing from the glass (although far less than shown here, only ca. 0.5mm).
  7. With an ohm-meter, try measuring if the nozzle is grounded indeed. (Edit: do this only when the printer is switched off, and disconnected from the mains supply, of course!) But glass itself is easily charged, so there is always a possibility that heat and airflow do charge the glass and the lower part of the print, even if the nozzle would be grounded. Or by pulling the print off the glass. This is guessing, but I think the main charge on models will come from post-processing: grinding, cutting off defects,... And it will depend a lot on the material. Do dust and small particles stick to your prints? Then they are charged. There do exist anti-static sprays: in the old days we used them on CRT displays. The glass of those big color monitors tended to get so charged up that you could pull sparks of several centimeters, causing sore arms, and burn-pits in the display, due to their 30000 Volt transformers. Maybe you can still find those sprays?
  8. Yes, exactly the same as I use; found it in a car accessory shop. :-) If you want to remove moisture from already bad filament, I think it is best to use both: a big bag of silica gel, in a hot environment (a sealed box on a heated printer bed, or in a computer-controlled oven). The elevated temperature might be necessary to dislodge the moisture from the filament. But be sure to stay below the glass transition temp of the filament, so it does not deform or melt. And then, once dry, store it in a closed box or bag with the silica gel.
  9. The purpose is to reduce inertia. Each gram less improves print quality, because you get less vibration, less "ringing" around corners, less overextrusion at corners (the nozzle can accelerate and decelerate faster), and you can print faster. A drive motor weights a lot, compared to the rest of the head. If the head is lighter, the gantry can also be made lighter, which further improves print quality.
  10. I would suggest you first practice your soldering skills on a broken harddisk or USB memory stick, or whatever else broken equipment with similar small chips. The main risks are that you damage the contacts on the board while desoldering, and next that you make short-circuits while soldering. Search on Youtube for tutorial videos, as there are a few tricks that can help. A video can say more than 1000 words. :-)
  11. I think the underlying cause of the printer-interruption you experienced, is that when a computer reboots, its USB-bus is always reset. But I am not a programmer, and I don't know if attached USB-equipment is supposed to be reset too? Or is it allowed to ignore the bus-reset? An argument pro resetting is that when a device is hung-up, it is reset and freed again, one of the purposes of rebooting. But for long 3D-prints, yes it can be inconvenient...
  12. This one has nice translucent colors. Maybe you can polish it a little bit, and brush off the white residu, but leave it unpainted?
  13. I have designed a compact spoolholder for myself, but it only fits one standard spool (Ultimaker, colorFabb). So, in its current form it won't work for you. But you are free to re-use the basic concept, and adapt it to your needs: you will have to make it bigger, and stronger, so it can support the extra weight. It uses one standard 608 bearing (=same as in most skate boards), which is centered. Keep in mind that frictionless spool holders also have a disadvantage: the spool may unwind by itself, especially near the end, if wound too tight. So you may need to provide extra guards to prevent this. This design is made in DesignSpark Mechanical (=a limited version of the commercial SpaceClaim, distributed by RS-components, and free after registration). If you have this software, you can edit the design. See here for the concept and files (and then scroll down a bit). https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ Purple = 608 bearing; blue = bolt and nut; cyan = base attached to UM-printer; yellow, lightgreen, red = rotating spool holder.
  14. If it is a "brand new" printer from 2017(?), maybe the filament has just gone bad? It is well known that PLA gets harder and more brittle over time due to internal crystal changes, and PVA absorbs water and then may fail too. Or maybe there is something else you overlooked or were unfamiliar with, for example some settings? Or there might be a manufacturing defect indeed? I would suggest you do a test print, and show *exactly* which settings you used, and show what the results were, both while printing and afterwards, with good close-up photos, and with all the files (STL, project,...). Then maybe people can already diagnose it. It could be something very trivial, hopefully. I have made 1000+ good prints with the previous model (UM2), and I have seen lots of very good models made with UM3 printers. So it should work well, or at least can be made to work well. But settings and characteristics differ a lot from liquid light-curing resin printers. Might be part of the learning curve. We all needed to do several prints before we got the hang of it, and it can be frustrating indeed in the beginning. You should be able to get results similar like these, without problems:
  15. It looks like the even and uneven layers consistently are shifted half a mm? But both are still present, it seems. Are both nozzles seated correctly in their holders, without any play? Or is the switching mechanism causing the issue? Maybe you can see if there is any play, if you watch closely while making a small test print? But I don't have a dual nozzle printer, so I am just guessing.
  16. If you look at the layer-view in Cura, I suppose it shows support in those areas? If so, it looks like there was no support at all in the areas where the spaghetti occured? Maybe the PVA-nozzle clogged and stopped extruding PVA, or there was huge underextrusion? (Too wet PVA?) But I don't have a printer with dual extrusion, so I can't give much suggestions for handling.
  17. Yes that is going to have an effect. The bigger the contact area of the bottom layer, the better it is going to stick. And the smaller the design elements, the easier they can bridge gaps without turning into spaghetti. I don't know the official terminology, but usually I speak about "raised" or "recessed" text or design features. So if the recessed seems to work better, definitely try that.
  18. The purpose of printing via SD card is to completely separate your printer from your computer, so they can't influence each other. Whatever you do on the computer, can't have any effect on the printer, as there is no connection. Or maybe you forgot to disconnect the USB-cable?
  19. In your photos, the next layers have nothing to stick to, so they print in the air and are dragged around. You need to find a way to make them stick. It is like when the glass bed is calibrated wrong, and you have too much distance between nozzle and bed. What about trying this method: - make sure your nozzle is rather close to the glass bed, so the first layer is squished very well into the glass, - make the indentations 0.11mm deep in the CAD design, - make the first layer 0.10mm thick in Cura, - make the following layers 0.2 or 0.3mm thick in Cura, Then the first layer is really squeezed into the glass, and there is a good change that the following layers will still have enough bonding. I am not sure this is going to work, and if the pattern is going to be visible enough, but it might be worth trying? This is what the bottom of the first layer should look like:
  20. What about printing this model flat on its back? Then there are no bridges, and no drooping spaghetti? Just rotate it 90° in Cura. Otherwise, if you do not use support, then on the highest point of the bridge, the filament will allways sag because a molten plastic (liquid) can not keep its shape without support. It needs a base to glue itself to, such as a solid previous layer.
  21. Maybe this behaviour is because sometimes the outer wall encloses multiple inner walls at once? For example when printing a lot of small circles touching each other. Then the inner walls may all be separate circles, but the outer wall might enclose them all at once. So it is just one uninterrupted outline. Just guessing, but I could see this as a reason. Edit: or vice-versa, as in this picture below: here the outer walls around the holes are separate circles, but the inner wall is one long connected outline.
  22. I have used this concept before when printing large overhangs. It was based on an idea of another user, but I forgot his name (might have been smartavionics?). It works very well, even with only a few connecting strands. My strands were 1mm long, 0.5mm wide, and 0.2mm high (=2 layers of 0.1mm). Inverted triangles to generate support do not work well for small layer-heights of 0.1mm: they tend to curl up too much. Inverted staircases (1mm steps) worked better for me. See the pics. So, depending on the design, this method can be very usefull. Inverted triangles curl up too much. Basic concept: Long bridge with hanging support, so the text below does not get damaged by the supports. All plates are 1mm thick in this test. The little ribs on top of the support are 0.5mm wide, separated 1mm. The result. The supports can easily be removed, and they do very little damage to the walls.
  23. I tried doing this manually: cut both ends at 90° angles, hold them together in a custom device (see pic below), heat a knife in a flame, put the hot knife inbetween both filament ends and melt them, remove knife, push both molten ends together, and let cool. Then you need to grind away the flange at the seam, otherwise it will not pass through the bowden tube and nozzle. You need to melt both ends to get a good bonding, then it is almost as strong as new. This method works and can especially be usefull for artistic purposes: to melt lots of different colors together. But it is not worth the time and hassle for me. So I use the left-over ends for doing atomic pulls, or for other purposes where I need a bit of plastic. For example you can heat a left-over strand, and ply it around something else (think of cable binders). Or ply them into hooks or clamps, or whatever.
  24. Maybe there are different infill patterns, or infill percentages, in which the nozzle follows a different traject with less starts and stops, and less jumps? I don't know if this is possible, but just guessing. Check this in layer view in Cura, before printing.
  25. It appears that OpenGL is coming with the graphics drivers. OpenGL is a specification to which drivers and hardware have to adhere, it is not a driver by itself. I am just echoing what I read on internet here (and hope it is correct), I am not a programmer. :-) So the latest graphics driver should give the latest version, if the hardware supports it. And if the manufacturer delivers new drivers for your system. Search for: "windows how to update opengl". Also there might be incompatible or buggy drivers, so in such cases you might need to try a different version, higher or lower than your current version. I am not sure, but I vaguely remember something that this was the case with some Intel laptop drivers?
×
×
  • Create New...

Important Information

Welcome to the Ultimaker Community of 3D printing experts. Visit the following links to read more about our Terms of Use or our Privacy Policy. Thank you!