geert_2

I don't know the official answer, but just from general computer experience: if this is on a Windows computer, could it be a filetype association issue? Try rightclicking on the obj file and "open with...", and then chose your preferred slicer. Or else try changing the filetype associations. Google for how to do this, as the method may differ in every Windows-version and language, especially on newer Windows 8...10.

Ultimaker has the new hard "breakaway" support material, maybe that could work? Also, another option is to use ABS itself as support for ABS, by designing the supports as part of the model, in the old way like for single nozzle printers. Concerning boxes: if you like to design one yourself just for the fun of it, or for a limited space, you could look for food boxes or freezer boxes to start from. They do exist in a lot of different sizes. Although I think the real cost is going to be more than the Polymaker box, with less functionality probably.

I have also noticed that when the nozzle is just waiting outside of the model, this may cause little defects due to the nozzle leaking, or due to its "take off" from the model, and due to the no-flow in the nozzle, which causes higher temps and thus more liquid material. Also sudden changes in layer area show up due to huge differences in cooling time. That is why I usually print a dummy model next to the real model, when printing time is too little for good layer cooling. Instead of using the minimum layer time feature. Ideally, the dummy model would have the inverse layer area of the real model, so cooling time per layer is identical. This doesn't matter for big objects, but it does for very small ones. So what you could do is subtract this model from a cylinder, and use that inverse cylinder as dummy? Be sure to design a good bottom plate to get good bonding to the print bed (for example: something similar to a brim, but manually designed into the model). For best results on small objects, layer printing time and material flow through the nozzle should be steady through the whole printing process. Another option might be to print enough models at the same time.

Might also be a combination of speed and hitting something, like a curled-up edge of an overhang, which caused it to miss a step?

An advantage of open source models is that you can buy components from other manufacturers too, if it would ever be necessary. This was one of our main concerns a few years ago and an important reason for selecting the Ultimaker2, instead of a closed source model of comparable quality. Because this market is still very young and volatile (it was especially 2 years ago). And indeed, some of our collegues also had bad experiences with customer service of very expensive closed source printers (60000 euro...), in Europe, so I understand your concern. Whatever you buy, I would suggest: make sure the printer is reasonably open source in case of hardware defects, and make sure it uses open standard spools of filament, (or a standard resin for resin-printers), so you can buy filament from multiple sources. Make sure it can handle multiple materials, like PLA, PET, polyesters, etc... And select a reliable brand with good reputation. If you have engineering-schools or fab-labs around you, they might also be a good source of information.

I have used PET (or is that PETG?) from the brand ICE. This has an indicated temp range of: 215°C...250°C. I found that it prints well at 220 to 225°C. Lower temp causes underextrusion, higher causes too much stringing and hairs. Layer bonding is quite good when printing slow: 25...30mm/s, if using no fans. I use no brims, no raft. Heated bed: ca. 80...90°C. At 70...80°C it comes off and warps, so 10°C more made a huge difference in sticking. To prevent warping, I print with fans off, on bare glass. No treatment of the glass, except from cleaning well (no soap, no detergents, only pure water). But printing without fans causes poor overhangs, and it has difficulty closing gaps on top of models, like a roof. When I need to print good overhangs, I need fans for cooling. Then I need glue on the glass bed (gr5's method: 10% white wood glue in water). But this chipped the glass once: while cooling, I heard weird cracking sounds, and when pulling off the model, part of the glass bed came with it. I think it already chipped the glass during cooling, at least partially, not as a result of moderate pulling. So, as long as I don't need much overhangs, I prefer to print on bare glass, no bonding, bed=90°c, nozzle=220...225°C, speed=25...30mm/S, no fans. Try a small test model first, a solid bar (e.g. 10mm wide x 10mm high x 50mm long) with some overhangs and sharp corners, for example, because these cause high warping forces. If that works well, other models should work well too.

3D-printed PLA may be more sturdy than ABS, because it has a much better layer bonding. It is less likely to split suddenly. I am often surprised at how much abuse my PLA models can have. Under load, they often survive longer than identical PET or NGEN models. But PLA has a couple of problems too, in my experience: - In a hot car, or even in direct sunlight, it *will warp*, no doubt. So keep it out of the sun and out of hot cars and similar areas. Unless you use high-temp PLA. - After a year or so, PLA gets harder and more brittle. It can still withstand quite some abuse, but it will no longer flex, and instead break suddenly. For example with snap-fit lockings, this is a problem, because you keep pushing hard until it locks or unlocks, and then it breaks. Thus you can not rely on flexibility to absorb sudden loads, and you can not design fragile springs, rather you need to design it like a Flintstone car model in concrete. - Under a continuous high load, PLA suffers from creep deformation and micro-cracks. Just like most other hard plastics. So, make sure you store the model in a no-load condition, e.g. the wings supported by foam. - You can not make a screw thread in it, it melts instead. Even very slow and light drilling is a problem. I just design holes in the models, and use nylon screws and nuts for clamping. - I don't know about UV-stability, but most plastics deteriorate quite fast in UV-light. I guess it will be the same with PLA. - Water is no problem, contrary to expectations based on the so-called bio-degradability of PLA. To test this, I have made a sift that is sitting for more than a year in my laboratory sink, and it still is okay. I also made one for colleagues, and theirs has been contaminated with blood and chemicals, and it is still okay also. So, your plane won't dissolve when flying in the rain, or landing in wet grass. I am also curious about the flying characteristics, since usually a "square" plane (=about as long as wide) and a traditional layout (=with a big wing for lift in center of gravity, and small control surfaces far behind, like in sports planes), are considered to give the best control and flying characteristics, at normal subsonic speeds. So I wonder how a flying wing compares to that? In what aspects it would be better, and in what worse? I would guess lift is going to be better, but yaw-stability is going to be worse? But what about drag and maneuvrability, and stall characteristics?

This sort of holes can easily be done in CAD: design one hole, and then repeat that in a pattern in both directions. This sift (for a laboratory sink) is ca. 50mm diameter. But it does not print easily: too much jumping and retracting. Making a solid plate and printing it with ca. 70% to 80% infill (try it) like gr5 says, would give much smaller holes, and will be quite strong, but it will have a *very high* resistance to airflow. So it all depends on the exact purpose. Maybe you would be better off printing a sturdy housing only, with big openings, and then put some standard proven air filter in it.

I haven't used CPE, but some PET I tried from a different brand, also had similar oozing and blobbing characteristics. These blobs came about due to the material sticking to the nozzle, accumulating, and then sagging onto the print. This caused brown blobs, and strings and hairs as the nozzle ran through it. Molten PLA is more like yoghurt, while that particualr PET was more sticky like halfway between honey and bubble-gum. Printing slower and cooler helped (caused less accumulation on the nozzle), but did not eliminate it.

That should work; I have done similar text and key chains with success. And without brims. Try my "salt method" for bonding PLA. Or try gr5's method (10% wood glue dilluted in water), or neotko's method (hair spray), depending on what you like/dislike most. As first layer, I use 0.2mm, and a good bed leveling, so this first layer is pushed well into the bed. For my salt method, see: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ Another option would be to make "watermark text", totally inside the model. Then you can have a 100% filled base layer without issues. Like this. Print this with the text towards the glass. But of course, this requires transparant or translucent material, and the top layer should be less than 0.5mm for best results.

For acetone smoothing ABS, I have seen videos on Youtube where people put kitchen tissue paper, or an old newspaper, all over the sides of the container. Then they poured in some acetone and let it soak into the paper. And then they hung the model in this container, without touching the walls. In this way, the acetone vapour comes from everywhere around the model, not just from the bottom of the jar. This gave a much more equal distribution, and thus better smoothing result, without need to cook the acetone (thus less risk of explosion, but be carefull anyway, since acetone is more explosive than fuel). Dipping ABS in acetone was not a good idea, if I remember well, since that dissolved the ABS too fast, and caused "tears" and deformation. For use on a car, you need ABS indeed, or another high-temp material. PLA will deform in summer, or even in spring or autumn on a nice day. Don't ask me how I know...

I can imagine this would fit very well in a modern minimalistic interior. It is an eye-catcher. I guess sooner or later someone is going to pick up this idea and commercialise it. So you might as well try and sell it first to the Expresso (or other) manufacturers.

Is there a way to increase the "auto-logout" time from the forum? I would prefer a logout-time of ca. 4 hours. Currently, it seems to be one hour? It still happens occasionally that I have to login again when posting. But at least, we don't lose the text anymore, after login-in again, it is redisplayed correctly. I don't use functions like "remember me" or "remember password" or similar, since it causes forgetting passwords, and it may cause security issues when you are logged-in automatically (not really important here, but may be on other sites, so it's a bad habbit).

What happens if you print this model, together with a simple test tower next to it? Thus a different design that you make yourself, and put it on the same build-plate? Do they both go wrong?

Indeed, if there is burnt residu in the nozzle, usually these flakes come loose when starting a new print, thus on the first layers only. I am not sure why that is, maybe because the cooling cycle between prints dislodges it? But I have often noticed that too. Sometimes I have to do 5 or 6 atomic pulls in a series before the dirt comes off (on my UM2, I don't have an UM3). Some materials work better for atomic pulls than others, for example Ultimaker Pearl PLA works well for me; it's a bit more sticky and glues better to the residu. Try whatever filament you have laying around. And before pulling, let it cool down long enough. If I print full-time, I do a series of atomic pulls once every week as routine maintenance (my gentle version, with rotating but without brute force). And more if necessary, for example when changing colors. Maybe you could also inspect the nozzle under a magnifying glass or microscope, to see if it is defective and leaks from halfway? This could happen on an UM2+ when the nozzle is not screwed-on tightly enough. I don't know if it can happen on an UM3, but if so, it could also explain these things.

The big pink/purple blob goes over *multiple* print lines, so it clearly is something that is dropped on the print from the *outside* of the nozzle. Problably some junk from a previous material that was still stuck on the outside. Might have been a bit stuck higher up, outside of direct view? Or some debris picked up by the nozzle while traveling around, then molten, and then deposited? Idem for the S-curved dark spot in the first pic. The tiny dark flakes along print lines, in the second pic, look more like burnt residu from the inside of the nozzle. This would probably require multiple atomic pulls to remove, and a lower printing temp (to prevent the material from burning). Printing temp is very close to burning temp, so you cannot totally avoid burnt residu, but you can reduce it by reducing printing temp and speed. Generally, in my experience (with two Ultimaker2 printers): big blobs, often light brown discolored = accumulated material on the outside of the nozzle that sags and then gets deposited on the model. Tiny, thin black flakes = burnt residu from the inside of the nozzle. To do more gentle atomic pulls, feel free to have a look at my manual (scroll down a bit on that page): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

Inbetween the time you last used your printer and now, did you install anything else? Or did you have any weird occurrances? It looks like something corrupted your graphics drivers and/or your operating system? Maybe some graphics software that installed corrupt graphics- or .NET drivers? Malware infection? Severe software crash with harddisk corruption? An automatic Windows update gone wrong completely, or interrupted? Or something else along that line? If *none* of your graphics software works anymore, it is on a system level, not application level.

Thanks. These are very interesting data, and a very good starting point for selecting materials. Maybe this thread should be sticky? Actually, I am a bit surprised at the positions of Taulman 618 nylon , and colorFabb XT and HT. A suggestion: as you accumulate data, maybe you could add a sort of "error bars" to the values in the second graph (with green, yellow, and orange ovals)? Both horizontal and vertical lines. These error bars would indicate the useful range. Then for example PLA might sit at 210°C, but its error bar might go from 190 to 220°C. While another material at 210°C might only go from 205 to 215°C. This would give an idea of the useful range of the material, and how critical temperature settings are.

I don't think this is going to work. The main problem is that most 3D-printers do cause hairs or strings, when jumping from one spot to another, especially for when printing very small pillars. And in this design they have to jump a lot. So you are going to have lots of short-circuits. If you would print thin full layers on top of each other (thus: layer1=pos, layer2=insulator, layer3=neg, layer4=insulator, layer5=pos, etc.) you might have more success, but still. When printing, the nozzle often gets dirty, and drags around material which is deposited somewhere else. This too could cause short-circuits. And due to the large distances between layers, you simply won't have enough surface to get a reasonable capacity. Finally, the high ohmic resistance (I don't know if this is the proper English term) is going to reduce the quality of the capacitor a lot. I do see the value of 3D-printed circuits for ohmic resistances, for example to power LEDs in model railroad trains or so, but for capacitors and coils, I don't think it going to work with this sort of printer, and least not in the near future.

What is the technical relation between speed and this bumpiness? I guess it has to do with higher pressure in the nozzle, but I don't see how a higher pressure would cause these irregularities? I would expect them near corners, where the head has to slow down, but not halfway a smooth curve? And if they would occur, why not on every layer, but seemingly random? There must be a valid technical reason, because the facts are what they are, but I just don't see it...

If you print a couple of these horseshoe clips, next time it becomes a lot easier to remove them and insert them. The basic dimensions are the same as the standard clips. I have been using these for 2 years without issues. For the design files, see my page (near the bottom): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

Or you could design the supports manually, as part of the model. Then you can use lines, ribs, thin walls, columns, blocks, or whatever shape you want as support. This gives you more control. And you can provide features to make removal easier, such as protrusions, or holes where you can insert tools. For example:

I think it is very unlikely that you run out of filament unexpectedly, since before the start of a print you can see how full the spool still is. After a couple of prints, you know how much your models will consume. The risks are rather: a blocked nozzle, filament grinding due to too much retractions, a kink in the filament that can't get through the feeder (I had this once), flexible filament getting stuck in the bowden tube, or something similar. None of these are "end of filament" conditions, so none would be detected by a simple mechanical switch. You would need something that detects motion: a laser mouse (standard mouse sensor only works for opaque materials, not for transparant materials), or a non-slipping rubber wheel and high resolution decoder driven by the filament (which also might not work reliable on smooth oily filament like Ultimaker nylon).

Another solution is to make the design in metric dimensions from the start. In the future, engineering in the USA and UK will switch to the international metric system anyway. They are about the only countries that are still behind. So if you do the switch now, you will have a very gradual and smooth transition, and you will be used to it by the time it becomes compulsary. Once you are familiar with the metric system, you won't regret it, because it is so much easier to do conversions. No more horrible multipliers to go from inch to feet to yard to miles to nautical miles... Just move the decimal point a couple of places, that's all. Idem for volumes and most other units.

I do not have an UM3. But a knocking sound seems like some play somewhere in the system. For example screws of the stepper motor coming loose, or something similar. Or something loose in the nozzle. Have you tried removing the bowden tube, but leave a little piece of filament in the feeder? And then manually move that piece of filament up and down? Next, idem for the nozzle: manually dial up temperature, manually feed a bit of filament, and manually retract it? Then Maybe you could feel where the problem is?