geert_2

Recently someone on this forum mentioned these links, about printing transparent models. Was that you @kmanstudios? I don't remember who. https://hackaday.com/2018/12/11/true-transparent-parts-from-a-desktop-3d-printer/?fbclid=IwAR3uP3W0fAE_ej6bp8VvflOeXPvFEv7yyHoH-jEt8jJBPWlaoNjq9E33tVg http://fenneclabs.net/index.php/2018/12/09/3d-printing-transparent-parts-using-fdm-fff-printer/ Contrary to popular recommendations, they used *thin* layers (plus slow and hot printing) to get good transparency. Might be worth experimenting with both unusually thin and thick layers, and comparing the results? Personally, when printing PET, I never got better than a "frosted glass" look. But printing very slow at 100% infill definitely improved results, even at lower temperatures, so the material has enough time to flow into all corners.

Keep in mind that the "salt method" only works for PLA on a heated glass bed. It does not work at all on a cold bed; there is no bonding at all. It does not seem to have much effect on PET and NGEN either (no positive, no negative), and does not work on ABS. I have no idea about other build-plate materials, such as aluminium. Also, I don't know the CR-10. So I would welcome feedback on your results. Be sure to stay around when trying these things, so you can abort the print if it would come off the bed.

Is he going to polish and paint it, or does it stay as-is? If used as-is, I also wonder what the effect of layer lines would be on the sound? Would that divert or filter some tones?

Mocht je het nog niet gedaan hebben, dan zou ik voorstellen om ook eens met leraars in het lager onderwijs of lager middelbaar te gaan praten. Vooral met degene die technische of wetenschappelijke vakken geven. Daar beginnen ze nu ook stilaan 3D-printers in te voeren. (In de hogere jaren zijn ze reeds redelijk ingeburgerd in technische richtingen.)

The bottom of your parts should look like this, when well squeezed into the glass. For PLA this surface is quite reflective: you can see the protective cover of a desktop fan mirrored in it. Top photo = focused on bottom surface of model. Bottom photo = focused on fan cover reflected in this surface. It is the same model, but I couldn't get both the bottom and the reflection in focus at the same time, since the fan was too far behind the model.

For tests I have melted filaments together: heat a knife in a flame, insert it between two filament ends (so they both get melted), and then push both ends together. I did this on a little tool I printed for this purpose (see picture). After that, the joint needs rework with a Dremel, to remove the ridge. This works well for small pieces of filament, but I don't see how it could be done on a whole spool while printing? Unless you are willing to do some acrobatics... :-)

I use a tiny injection needle, of which I cut off the sharp cutter tip, and rounded off the edges. Sharp edges might damage the soft copper or brass of the nozzle. But even then I take care not to apply a sideways load, so I don't wear out the circular nozzle-opening into an oval. Originally the needle was 0.41mm thick, so I sanded it down to 0.39mm, to make it fit into the 0.4mm nozzle opening. This is quite soft steel, and will bend rather than break.

In PET snap-fit lockings work well, if not flexed too much. But I would not use it for regular PLA: it gets too hard and brittle after a year, and then they tend to break. I have no experience with tough PLA. Depending on how secure the locking has to be, a screw and nut might also work. If you design a cage with some retention for the nut, so it can not fall out. Like the cages in these pics (although they are for a totally different clamping system, not directly applicable to a lid, but maybe you could use the basic idea).

I am glad it worked well. Normally I apply a few drops of very salt water on the glass (not too much, no flooding), using a pipette, and then I gently wipe it with a paper tissue until dry. It takes only 30 seconds or so. I found the salt method generally works very well for wide and low objects that can be printed in a couple of hours (as typical for my dental models). But it works far less for narrow and high objects like statues that need all day to print: then edges tend to curl up. So if you would ever try it for a high statue, be sure to stay around so you can abort if it should come off. Or apply a brim or Mickey mouse ears to the model. I still don't know the chemical *why* it works, since salt isn't exactly known as a glue. I guess it has to do with increased surface tension due to the salt, but that is only a guess. Concerning printing on fabric: maybe the stuff used for thin white curtains might also work? It is usually quite strong. Not sure what material that is: glass fiber, nylon,...?

Maybe you could reduce infill, if that would be what uses most material? If you need a certain wall thickness for strenght, set the walls to way more than the usual 2x nozzle width, lets say 20x nozzle width (e.g. 8mm instead of 0.4mm). And leave the rest empty or sparsely filled (20...30%)? I don't know what you are going to print, but for tires for lightweigh toys, it *might* work? For a heavy equipment tire, it won't. Or something along this line?

Probably you would have to do that in Pro/E itself I guess, before exporting to STL? But before you spend time doing all this modeling, you might want to try printing the texture on itself first. Load the JPG directly in Cura, set its height (I don't know how, never tried it, but it should be possible), and print that as a flat plate with relief. Chances are that due to the layer lines, and the width of the nozzle, the texture will get damaged to such a degree that the result is actually worse than without texture. Or that it causes other issues with retraction, or with stringing. It all depends on the texture itself, and the capabilities of the printer. Alternate solutions might be a few simple decorative ribs or a nice logo, or so.

The cube looks quite okay, I think. Could it be that the round object has too many triangles, thus way too many segments on each circumference? I vaguely remember that this has caused similar issues for other people in the pas?. I would suggest you design a new simple round button, and save that as STL with not too much segments. And then try again.

Maybe they just came loose from the build-plate, and started to wobble? Edges on overhangs tend to curl up, and then at the next pass, the nozzle might brutally bang into those curled-up edges. I am guessing here, but I have had something similar happen during overhang tests. See the pics: here they curled up ca. 1.5mm, for a layer height of only 0.1mm! Increasing the feet (the mickey mouse ears) and making them a little bit higher, should improve bonding. Printing cooler and in thicker layers, should reduce curling. At least, it did for me. At least, it allowed me to make nice spaghetti photos.

At first glance it looks like underextrusion, thus not a text problem but a general problem, as there are holes in the base plate too. But it is hard to say without more info (printer model, material, temp, speed, flow rate, etc...) You could try to print the same plate, but without any text at all, thus just a thin flat plate. At exactly the same settings, and see what comes out?

What if you put a piece of plastic or cardboard in the front opening, like a sort of primitive door, but with a few holes in it? So there is some circulation to evacuate the heat, but not too much? At least, cardboard is cheap, so it could be a good test to see if this really is the cause, and not something else that happened to occur at the same moment (e.g. dirty, oily glass).

In cold very dry winter weather when it is freezing outside, I have this too on my two UM2. But I also have it when touching the buttons of the elevator, the metal frame of the fume extraction cabinet, etc... Sometimes it causes sparks up to 10mm. We have a composite resin floor here, which seem to worsen the problem. In summer or in moist winter weather, there is no problem, thus it is clearly static charge. Touching the metal with the back of your hand is less painfull, since there are less nerves. Or touch it through clothes first, with your elbow, for example. Also, I have connected a *very high* ohm resistor to ground in a nearby wall socket. If I touch that first (if I don't forget), it discharges me. Maybe placing antistatic rubber mats on the floor, properly connected to earth, and using antistatic wrist wraps or similar, might also help? In het 1980s, we had big graphic supercomputers which also had that problem (and then the computer would reset and reboot). So, one customer solved that by watering the carpet with his plant watering can every morning. Not good for the carpet, but better for his business than losing his work every few minutes. :-)

I have no experience with flexible materials, so can't really help. But what if you disconnect the bowden tube at both ends, and manually try to push this filament through? And then reconnect it at the nozzle-end, heat up the nozzle, and try again, thus manually pushing from the back? Also, manually feed filament directly into the nozzle while it is hot. Maybe these tests might give you an indication and a feel for the amount of friction and resistance?

I like the fine and inventive engineering of this, with that geneva switching mechanism, leak guards, airflow directors, etc... If you could indeed make it more compact, it has the potential to become a hit, I think. Very well done.

Thanks for that link. Remarkable that they achieve best transparency by doing exactly the opposite of what is usually recommended: very thin instead of very thick layers. I am going to try that next time I print in transparent PET.

The grease is for the Z-screw only, not for the X- and Y-rods, nor the Z-rods. The Z-rods need no oil at all, if I understood well, since they have linear recirculating ball bearings inside (I don't know the proper English term). The X- and Y-rods need some light oil, like for bikes or sewing machines (but not dislodging oil, thus no WD-40 or similar). Although I use an hydraulic oil for fine hydraulic machines, because I have plenty of it in spare, and it does not dry out at all, and it lubricates well. I found that the sewing machine oil I first tried, dried out too soon into a sticky layer. But that doesn't explain vibrations at the rear side of the model only, when the model is sitting in front of the printer, indeed. Or do the vibrations only occur when traveling from left to right, and not from right to left (or vice-versa)? So there would only be stuttering in one direction? I would try to further diagnose by manually moving the nozzle, with the printer off.

I would not use a hair dryer: this tends to melt the filament, and then it shrinks in length but becomes much thicker in width. Similar to when you heat a piece of plastic above a flame (without it catching fire): it shrinks into a ball. This may actually worsen the problem. I would rather ply the kinks in the opposite direction by hand, very carefully. Or ply that bend in the opposite direction around a tube or something similar, to not overdo it. In my experience, it does not have to be perfect, but just good enough to not cause too much friction in the bowden tube and nozzle anymore.

Yes, this is an attractive model. I was also wondering: what is painted, and what is purely 3D-printed, like the gold for example?

Yes, this is a good advice. Make a compact test bloc with a lot of different screw threads in steps of 0.1mm or so. And print that in different materials, speeds, and temperatures.

Did you check if the rods are well oiled? With the printer off, if you move the head by hand, it should move reasonably smoothly in all directions. At least, if it is an Ultimaker; I don't know about other brands with other drive mechanisms. If it would only be a visual effect, without mechanical shaking (especially in an Ultimaker2), it could be caused by a tight bending radius of the filament (e.g. near the end of the spool), causing a lot of friction in the nozzle and bowden tube, and thus causing underextrusion?

Het zou kunnen dat de nozzle verstropt is. Dan moet je die proper maken met zogenaamde "atomic pulls" of cold pulls. De procedure kan je vinden op de Ultimaker site. Of je kan het evtl. ook proberen met mijn procedure. Zie hier (en scroll een beetje omlaag tot "gentle atomic pull"): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ Er kunnen ook nog andere oorzaken zijn, maar daarvoor google je best even op het forum.