Jump to content
Ultimaker Community of 3D Printing Experts

rewolff

Member
  • Content Count

    102
  • Joined

  • Last visited

Community Reputation

0 Neutral
  1. rewolff

    print is wider at the bottom.

    No pitch is the distance between the lines. The "first layer height" is set to 0.3mm as you say. As a compromise between "faster printing" and "better quality I've set the layer height for the rest of the print to 0.15. The "first layer width" is set to 1mm, it's not in the advanced settings but in the expert settings... Hmmm, I now see it is under the heading "RAFT" so it may not have anything to do with the "object-first layer".
  2. rewolff

    print is wider at the bottom.

    Yes, some production processes require manual or non-manual postprocessing. But the idea of a 3D printer is that I can make stuff that I wouldn't otherwise be able to. One of the things I want to be able to make is "gears". The "gear train" of commercial products are usually injection molded nylon. But there are so many different sizes and shapes that it is not possible to just have a few in stock and build something useful out of it. Lego and Fisher-technik come a long way to solving this. Anyway, if I print a gear and it has a base that is wider than the rest, it will not allow me to put the gears close enough to the other gear to have the teeth mesh over the full height. When there is just one gear meshing with another gear, you can turn one around and have the extra "rim" fall outside the other gear. So this way you can have about 80% of the gears meshing with the other one. But that won't always be possible. Anyway... We levelled the bed again, lowered it a bit. That helped. We lost about 50% of the "bulge" at the bottom. The gear went from 11.6mm at the base to 11.1 mm at the base, with a 10.6mm measurement at the "flat" part. We lowered the bed-temperature to 90 degrees (from 100). A print we tried then (with a big base of about 100x35mm) simply didn't stick. It came off after two layers. We increased the temp back to 100, and finished the print. -> 90 is too low, 100 is about right. (I'm not sure how accurate those numbers are in an absolute sense....) My suggestion to lay down the first layer with a wider line-width has already been implemented in cura. I was going to splice together the gear with a different line-width setting when I noticed that the first layer didn't change. The weird thing is that although the setting was "1mm", it seems to have a 0.8mm pitch in the print. The other layers are normally set at 0.4mm, and measure out to be very close to that. Does anybody know why the first-layer-pitch doesn't come out to the set value?
  3. rewolff

    print is wider at the bottom.

    You're saying to make my bottom face one line-size smaller. So both the first and second layer get to move into the leftover spot. Ok. That's another option. (but if I'm over-extruding, that might be causing problems on the inside somewhere as well.). The thing is: For a one-off project, just modifying the object is simplest and fastest. And it works to test out a theory like yours. But in the long run, I think we should define the models to be "the way we want the resulting object to be", and that the tool-chain is able to adjust for the inaccuracies of the building process. Suppose I build a few prototypes on my home-3D printer, and then go to order it at shapeways on their fused-nylon process. If I've made a bunch of adjustments in the model for my home 3D printer, then I'd have to undo those again before sending it off to shapeways. This is bound to go wrong at some point. We should work towards the tool-chain helping us getting as close as possible to the model. If you adjust the model for the printing process, you're going to be adjusting everything again the moment you move to another printer. (e.g.: I have two friends with a 3D printer where I can print things if I want to. Different brands....).
  4. rewolff

    print is wider at the bottom.

    I've been thinking about this some more. One of the problems of 3D-printing with extruded plastic is making the object stick to the platform. If you just extrude a strand of plastic it will just "lie there" on the platform, and it won't provide a base for the rest of the print. So probably to make it stick, I have a tendency to configure the software/hardware to provide say plastic for a 0.2mm first layer, while the hardware has only 0.15mm clearance between the head and the platform. This creates some extra pressure and makes the object stick better. But the "extra material" has to go somewhere. So for a few layers there is an over-extrusion that causes the bulge. But providing "just enough" plastic for the first layer makes it stick less... What to do? While, for resolution purposes you normally like to use settings like: "I Have a 0.4mm nozzle, so lay down lines of 0.4mm", for the first layer if you change that to: "Please lay down a 0.8mm line", you will extrude enough plastic for a 0.8mm line, which is extruded through the 0.4mm nozzle, pushes against the platform and the flat part of the nozzle before being left alone to solidify. I don't have dayly access to the 3D printer (I have friends with printers, but not one of my own). To test this without having to modify the slicer, I'd slice a print at "my nozzle is 0.8mm" and at 0.4, and the splice together the start of the 0.8 gcode-file with the rest of the 0.4 file after the first layer. Anyone game to try it? (in case you're not into messing with gcode files: you do the slicing, I do the splicing: send me the files) @gr5: the macro-foto brings out the "worst", besides the bulge at the bottom, it looks reasonable in real life.
  5. rewolff

    print is wider at the bottom.

    The print I made some 1.5 years ago was * PLA * no heated bed (and I have that in front of me). The print I made a few days ago (and my friend's) * ABS * heated bed at 100deg. WITH capton on the bed. This is the PLA thing that I now have access to: The effect is more pronounced with the ABS.
  6. Hi, I'm trying to print a gear. A friend is making an object that ends at say 60-70 degrees to the platform (i.e. on a small base, growing slightly bigger upwards). My friend's object is getting printed and the print starts out with a vertical part. The base is too big. Similarly my gears (vertical walls) are sloping inwards for the first few mm. The rest is fine. Is there some obvious setting we have wrong? Is this unavoidable?
  7. I filed a bugreport on github yesterday mentioning one of the named items here. Daid pointed me here. So while I was on my way here I thought about "what else can be improved?"... Everything I came up with has already been mentioned. (plus a few other good ones). Thinking some more... Why does cura visualize the models? To place and manipulate them! Why visualize the toolpath? Now it's just "check". but in the (far?) future it might become a toolpath editor. So, within a layer there are a bunch of infills and a bunch of moves. I might want to change the order of the infills between the moves. Similarly, a friend's printer has the tendency to have the overhangs "bend up". So a move that comes at an overhang will bump into the overhang and possibly bump the object off the bed. So "editing" a move (in this case) should be possible: Change to: "move up 1mm, <original move> move down 1mm" or: take the long way around: do not pass the overhang. Or: "added constraint: "Use this move in reverse direction and optimize layer again". Sometimes the layer-changes are important. So a "please start this layer here" would also be an operation that the "gcode editor" could allow.
  8. rewolff

    Cura 14.05 RC2 - bridging

    A friend has a new 3D printer. When I played with an Ultimaker a long time ago, the PLA was "easier" and I didn't see this problem. However, with the ABS shrinking a bit after solidifying, the overhangs tend to bend up. My friend doesn't have anything but ABS yet, so he's seeing this problem: The head hits the overhang, and potentially bumps the object from the bed. Solutions: I don't know. I'd think that doing the overhang "too high" when there isn't anything below, could work. Then approaching the material on the other side from the top might help. But difficult to parametrize: how much up is enough? How slow do you need to make "XY" movement for the Z to "catch up". And even if you get it right, maybe you're creating a bigger bump, causing even more problems on the next layer.... Hard!
  9. Currently things on the software side are already limiting. For example, the general idea is that every 100us, it is evaluated weather or not another step is required.... So at normal speeds, sometimes yes, sometimes no. This already results in artefacts. If the Y axis is taking a step every 3 cycles, but the X axis every 3.1 cycles, a visible pattern can be seen. The 100us was chosen because the AVR in the arduino can handle it, but not much more. Fine. But having a 5x faster CPU can push the artefacts a factor of five further down the line. Keeping things whole means sensible limits for the hardware have to be set. Max acceleration thisway and thatway. The software simply has to obey the limits.
  10. rewolff

    conection problems!!

    Oh. I remember that "device cannot start". This happened when the Prolific PL2303 windows driver thought it was talking to a fake PL2303 chip. They were annoyed at so many fake chips doing the rounds so they modified the driver to give that error message when it detects a fake chip......
  11. rewolff

    Walls are too big

    The above post is SPAM trying to promote their website, hoping to trick moderators into leaving it there with the formulation of the message like that. I was tricked until I couldn't find any link....
  12. rewolff

    Filament without spool

    I just put it on the floor behind the printer. Make sure to uncurl the spiral between the stock and the printer every now and then. You can have a big "spring" like structure between the stock and the printer so that you can leave it alone for hours on end....
  13. IMHO, going from "hot enough to melt the plastic" to "plastic stays solid" should not happen say inside the brass. Suppose you've been printing for quite a while with little extrusion (lots of jumps). Now everything heats up, lots of plastic melts. Next the print is finished and you shut it down for a day. Next day you try to print something.. but it may take a very long while to melt the plastic as far "up" in the brass so that it un-sticks. What is needed is a "sudden" transition from where it melts to where it doesn't melt. This is accomplished by the changeover from brass to peek. The brass is hot, the peek should be cold (below plastic-melting-temperature). Next, some of the molten plastic will work its way up the peek tube (or whatever you're replacing it with). With the same scenario as before, you want some extrusion the next day.... So, what we need is that the solidified plug in the peek is simply pushed down into the hot zone, and melts there. This means that the peek or teflon needs a very low friction coefficient. This allows the plug to be pushed down as normal.
  14. Lots of plastics "remember" what happened around the time they cooled. So you have those food-containers that suddenly "remember" being flat if you heat them to 60 degrees. There is a bit of "tension" in printed PLA. even though it's less than ABS, I've had trouble that a larger box came loose from the bed due to shrinking. So what PLA "remembers" when you heat it is that it was supposed to be a bit smaller, but it HAD to be bigger to fit on the layer below. However the neatly extruded 3mm filament doesn't remember anything. Maybe the odd-looking curved sections have something to do with that. But the diameter stays firmly the same....
  15. rewolff

    What are shells and how to create them?

    I have a test-object where we print 0.2 mm through 1.6 mm cylinders with 0.2 steps. The 1mm one always comes out more or less as two separate cylinders, an inner one and an outer one.
×

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!