geert_2

If you would have a Dremel or similar tool, and a cutting disk, you can cut screws with that disk.

That are overhangs, which means they will need support material to print well, or they will sag a bit while printing if you switch off support. Has nothing to do with X-axis. See an extreme example in the testmodel below: since there is nothing to support the long overhanging arc, the printer prints in the air, and the extruded sausages do sag. As you can see, the problem starts at a certain angle of overhangs. This angle depends on layer-height, material, temperature. If I would not want this sagging, I would have to switch-on support. Then it would print a sort of construction scaffold below the overhang, to support it. If you are totally new to 3D-printing, I would recommend that you design or find a couple of test models, and print them with various settings. Keep watching closely while it is printing, so you see exactly what happens. In this way, you will get a lot of understanding and a feel of it, in a short time. It is well-spent time.

If it is crashing randomly in every version, then you might have a hardware problem: memory chips or connections that have gone bad. Sometimes heavy programs with complex graphic functions do use memory areas or CPU-functions that simple programs never use because they don't need that memory or functions, so only the heavy programs will crash. A bad graphics driver or card could also be a cause, because graphics cards do a lot of the 3D-calculations. Software bugs would normally crash in the same way at the same point, if you do the exact same thing. Like in a very old Word-version: if you had spelling checking on, and then you would change font color, it would crash, very repeatably. If spelling checking was off, it never crashed when changing font colors. That is more the behaviour of bugs. What sometimes helps, is opening the computer (after shutting down and unplugging from mains), and then with compressed air blow all dust away carefully, and wiggle a bit with all memory cards and graphics cards. So the connections are refreshed and oxidation is removed. Be carefull not to damage anything. Do this only if you are familiar with computers. Then switch on and try again. Also verify if your graphics card has a fan, and if that fan is running smoothly. Also check the CPU fan. Using analysing tools, check if chip temperatures of CPU and graphic chips are not getting too hot: a bad fan could also cause such problems. There is no guarantee that this will work, but sometimes it does if the crashes are due to bad or oxidated contacts, bad fans, or false connections due to dust. It will not work if a chip is damaged internally, or in case of a bad driver, of course. This does not mean that there can't be a bug that affects all Cura-versions, of course. It is just the random aspect that is often a result of hardware problems, rather than software. PS: I am not related to Ultimaker, just a user.

Except for cameras, microscopes and similar special equipment, people rarely use fine metric threads, because the coarse thread is already quite fine. So any "M3" should be a standard coarse M3 thread, if not explicitly indicated otherwise. If it is a fine thread, then it would say: "M3 Fine", or "M3 F", "MF3" and/or give the pitch "M3 x 0.35mm" or "M3-0.35", or something like that. I don't know what the official notation is, but I have seen all of these. Not to be confused with "M3 x 20mm" in which 20mm is the length of the screw (including head in countersunk screws, without head in standard heads). If no pitch is mentioned, standard is M3 with 0.5mm pitch. Can't you find those M3x4 screws in a hobby hardware shop? I think it is a quite common size. Or else, maybe take a longer M3 screw and cut off a bit?

Maybe the power supply gone bad, or degraded, so it can no longer deliver full output? Or bad contacts? These are the first things I can think of. To be sure, also check the mains power, if that voltage is correct. Should be okay, but you never know...

That seems to be underextrusion, but I don't know your printer, so it is hard to say why? Filament that is stuck on the spool or entangled, or bent, so it can not easily unwind and pass through the nozzle? Or irregular diameter filament, too thick to pass through easily? Or hard parts in the filament, that temporarily block the flow through the nozzle? A partially clogged nozzle? Something along that line of thought maybe?

That coupler definitely needs replacement. It is hard to see on photo, but it looks like its outside has swollen too? Normally it is only the inside that gets deformed by the pressure and temp of the molten plastic, and the contact surface that gets burnt from contact to the hot end. I think you should have a closer look at the fan again? Nothing blocking the air flow? No piece of plastic or other obstruction blocking the air flow? You should feel some flow. Else, also try to check the nozzle temperature? Is the nozzle temp okay, not too hot? For example if the sensor would be sitting too loose and not make good contact with the block? (Be carefull, don't pull on the sensor wires, they might break.)

In the UM2, the sensors do tend to get stuck in the nozzle heater block assembly indeed. I have been wondering about that question too. In the chemical industry all bolts and nuts got some anti-seize paste (don't remember the formula, but its color looked like dull silver); and in car- and ship-mechanics, people tend to use copper grease on all steel bolts. This not only lubricates, but prevents corrosion too. So maybe copper paste would do? At least, it should give good thermal contact. But I don't know if the grease would dry out or burn out at 260°C, and maybe cause more friction than using nothing at all? As for the white computer heat paste: that paste often seems to dry out and get hard and sticky, or it crumbles, after a couple of years. So I am not sure if that would be a good idea? Also, I don't know if it can withstand 260°C? Computer chips generally should not get much hotter than hand-warm, if you want some decent life out of it. Except for a few high-power amplifiers that can handle higher temperatures. So you would really need to make sure the paste can handle those high temperatures for a long time.

I think of it like a paint brush: if you try to paint strokes twice as thick as the brush (in one single pass), it is going to look irregularly and messy, even if the brush would be followed by a sort of scraper to flatten and spread the paint. Idem if you try to print strokes only half as wide as the brush: they will be very irregularly too. You will get the best results when the paint strokes are as wide as the paint brush. The molten plastic is a thick liquid. By increasing pressure or reducing speed, you might be able to squirt out a lot more liquid than usual, but it is going to flow in all directions, out of control. Liquid tends to flow into the path of least resistance, and once a flow is going into one direction, the rest behind it easily follows, creating a stream, but leaving the rest of the area uncovered. And vice-versa, if you want to print too thin lines, the flow is going to stutter and produce irregular lines and blobs too, like a thin leak that drips, thus in an on-off-on-off-... flow. Similarly, if you would try to print layer-heights of 0.8mm out of a 0.4mm nozzle, it might less or more work if the molten plastic would not leak away too much, but it is going to look very irregularly and messy. I once tried 0.4mm and that already looked poor.

Andere optie: probeer DesignSpark Mechanical (DSM), dat is ook gratis en heeft bij mij nog nooit enig probleem gegeven bij het maken van STL-files, in de setting "fine". DSM maakt perfect waterdichte, solide modellen voor 3D-printen. Het heeft ongeveer hetzelfde push-pull concept als SketchUp, dus je kan het heel snel leren. Zoek op Youtube eens naar tutorials en demo's, er zijn een massa tutorials van te vinden. SketchUp is ontworpen om visuele 3D-modellen te maken, zoals in computerspelletjes. Het is niet geschikt voor 3D-printen: het maakt een soort "kartonnen" modellen, waarbij de wanden niet goed aan elkaar passen en er altijd spleten in zitten. Dat maakt het model ongeldig en onprintbaar, want het is niet waterdicht, en niet solide: de inhoud is leeg met eindeloos dunne kartonnen wanden, inplaats van massief gevuld te zijn. Als je met SketchUp blijft werken, ga je eindeloze frustraties en tijdverlies tegemoet, het zal nooit werken. Stap daar ineens vanaf en probeer één van de vele gratis alternatieven. Dat kost je een paar weken inwerktijd, maar die haal je er heel snel terug uit.

Hoi, Je kan beter in het Engelse deel van het forum posten, daar zijn veel meer lezers en krijg je meer antwoorden. Als er niets uit de nozzle komt, zit die ofwel verstropt, of hij warmt niet op, of er is iets mis met de aandrijving, vb. het feeder weel slipt, of het filament zit in de knoop of te vast op de spoel, of de ventilator in de printkop werkt niet waardoor de hitte omhoog kruipt in het filament en de toevoerbuis verstropt, of de kop zit te vast tegen het glas, of zoiets. Kan echt vanalles zijn. Dat geeft allemaal underextrusion. Op één van de forums staat een zeer uitgebreide handleiding hoe je de oorzaken van zulke underextrusion moet onderzoeken en oplossen, met talloze tips. Ik denk van user "gr5", maar ben niet zeker. Ik denk dat user "AndersK" ook eentje heeft? Alleen weet ik de links niet, dus je gaat wat moeten zoeken. Zoek evtl. ook gewoon op internet naar: how to solve 3d printer underextrusion (en soortgelijke termen, al dan niet met je printermodel erbij)

Another way to do it would be in CAD: place all parts in the correct position in CAD, export to STL, and it will print as one design, all in the right place.

To me it looks like warping too, due to poor bed-adhesion in those areas. Make sure there is no dirt and no oil/grease on the glass, and no soap. Clean with whatever you want, glass cleaner, alcohol,... Then clean again, only with isopropanol alcohol. Then with warm tap water only, absolutely no soap (=reduces bonding), no additives. Especially if you print on bare glass without bonding. Then use any bonding method you prefer, if any.

It also depends on how flexible it has to be, thus how resistant to warping? When I print carabiner hooks in PLA, they are much harder than in PET. However, I need to bend the carabiners to make them hook around stuff, and then the PLA carabiners break sooner than the PET. PET is a bit more flexible, so it can withstand more warping. But after a year or so, they also tend to fail. PLA generally cracks before it fails. My PET generally snaps all of a sudden without warning. Could be different for different brands. Also, PLA can be glued easily with cyano-acrylate glue, after roughening its surface with sandpaper or a file. PET is much harder to glue. If it has to sit in the sun outdoors, laying around on sand or stones, PLA is going to warp for sure. You would need a high-temp version PLA (+60-70°C). PET can withstand the sun, at least here in Belgium, Europe, in our moderate climate. I don't know about deserts. If it has to snap-lock around things, a PLA lock will initially snap much harder. But due to creep deformation, it may loose its clamping force. And it may become more brittle and break after a year or so. For printing, I would recommend printing very slow, so the new melt has time to warm-up the previous layer and bond well, and the extruded saugages have time to spread into all corners. Also print in thin layers, that gives less indents, less layer lines. But printing slow makes the filament sit in the nozzle for a very long time, so reduce temperature to the absolute minimum of the range of that material, or even slightly below the range. So it does not decompose in the nozzle, nor clog it. I would print at 100% infill, and a couple of outer lines (e.g. 2x or 3x nozzle width), to give the best interconnection and least amount of internal gaps that could cause stress. But some people say that a bit less infill would give more flexibility and more ability to absorb shocks. I don't know. So, that might be worth comparing too? Make a small test model, and do the comparative tests on it. In general, if it was for myself, I think I would rather go for PET. My PET-prints do survive in the car, PLA-prints don't. If not strong enough, then I just make them thicker... If you have nylon or polycarbonate, or maybe PU, they could be way stronger and/or more resistant to breaking, but I have no experience with printing these. Seems to be more difficult. Carabiner hooks: left in PLA (cream), right in PET (green) Note the cracks and permanent deformation in the PLA parts, they are close to failure: The model in CAD: The fracture-surface of a PET one. Notice how the fracture-lines go diagonally through all layer-lines, they don't follow the layers, or not too much. You can see the primary and secondary starting points, the centers of the star-patterns. This one snapped very suddenly, without prior cracks or warning, typical for this PET:

Also, from the very beginning keep in mind that 3D-prints have lots of indents and pores where dirt accumulates easily, due to their layer lines and extruded "sausages". You need to consider smooth printability too, to minimize these. Also consider a material that can be post-processed and polished well, to remove most of the roughness. Not all plastics can be polished well, some are too fibrous and get ragged. How sterile does it have to be: only "free from living organisms", or also free from any accumulation of dirt and dead organisms? Those non-living accumulations could be toxic for your cultures, or they could be undesired food (proteins, fats, hydrocarbons), or they could just change things. Heat won't remove the "dead bodies". But any agressive chemicals and solvents that would, would also attack the plastic, and/or seep into it. And thus later-on evaporate from the plastic and destroy or alter your cultures. If UV-light might also be an option for desinfecting, then you need UV-resistant plastic. And preferably UV-transparent too. So I would at least go for transparent plastic, if possible. Then you can see if there is anything inside, or entrapped air. Avoid ethylene-oxide (etox) for desinfecting: it is extremely dangerous: causes severe chemical burns that you don't feel until 24h laters, when your tissue is dead and needs to be amputated. It is cancerous. And it is *extremely* explosive: it exothermically reacts with almost everything, decomposes and explodes. And it can explode in concentrations from 3% to 100%, thus almost always. The spark from dropping a metal tool is enough to make it explode. For the least amount of gaps, entrapped air, and ridges in your models, print as slow as you can, at or below the minimum temp for the material (otherwise it decomposes in the nozzle due to sitting there for too long), and at the thinnest possible layers. See these tests: layer-heigt from left to right (mm): 0.4; 0.3; 0.2; 0.1; 0.06, all from a 0.4mm nozzle. Printing speeds: top-row 50mm/s, bottom row: 10mm/s. Text caps-height is 3.5mm, leg width is 0.5mm, depth is 1.0mm. The model: 20mm x 10mm x 10mm, with hollow watermark halfway Best-quality block (0.06mm layers, 10mm/s): left as printed, right after polishing a bit.

Yes, you really need a hardened nozzle for glow-in-the-dark: it seems it contains very hard crystals, harder than brass. But I don't know which type of nozzle is suitable, no experience with that. Maybe hardened steel might do, maybe not and it needs to be a ruby? Try finding the hardness of the particles, and nozzles?

That is going to be a hard one. Even if you could model it, an FDM printer is not going to like it. I have printed a flat plate sift, not curved like this, by playing with the infill-parameters: print only outer side walls, no bottom, no top layer, and then set the infill to maybe 70%. Then study the preview in Cura, and adjust as desired. This was achieved by playing with Cura-settings, not in CAD. But I don't know if there are any infill patterns suitable for this complex curved thing? My holes only had to go top-down, not sideways. For modeling in CAD, I once created a simple flat pattern of holes, and extrapolated that over the model. But that was a *very simple* flat thing too. So I don't know if you could extrapolate and map a pattern of holes over a complex curve? But that is the basic idea: - make a hole in a plate - extrapolate that hole to fill the whole plate with holes - then map that pattern over your model, like you would map a texture over a model A very simple hole-pattern extrapolated over the bottom of this sift in CAD. Then edited as required. A crude filter (designed as solid, without pores) printed without bottom layer: thus side walls were solid, but the bottom was only infill, so that bottom became porous due to the infill pattern. This was only for capturing chunks of dirt, not for fine dust, obviously.

If it is only for one print (e.g. to test it), you could manually edit the gcode-file in Notepad, Notepad++, or a similar plain-text editor. But be aware that if you insert wrong codes, you could make the head crashing into things and damage it. Or you could burn the plastic in the nozzle by heating it too hot for too long.

I think the easiest, cheapest and fastest solution would be to go to a brico-shop, section "cart wheels". Wheels with a 100mm diameter start from around 10 euro. They will be much more durable than 3D-printed wheels, and much cheaper. You could use cart wheels, remove the outer metal holder, and only keep the wheel. In brico-shops, they also sell metal rods for making the axis. Hubo has a huge selection of wheels: https://www.hubo.be/nl/a/bouwmaterialen/transportmiddelen/wielen-velgen.html

I haven't experimented with ironing yet, so I can't give advice on that. Concerning fighter jet cockpits: long ago, maybe 40-50 years, here in Belgium and the Netherlands you could buy a real fighter jet for the price of its scrap metal. Weapons and engine were removed, but apart from that they were often intact, complete with all avionics in the cockpit. Just worn-out from 20-30 years of use, of course. It were F-104 Starfighters at that time. So several people had one in their garden as a toy, and even some kindergarten playgrounds had one. So the kids could play pilot. At that time, we considered that "just normal". Just like an old abandoned car in the garden... Today I know one remaining in the south of the Netherlands, not too far from the city of Zierikzee. I wonder if that would still be possible? If you just call the military, and ask to buy the whole cockpit section of an abandoned airplane from their scrap yard, or the whole plane if you have enough room? Would be a good starting point for making your own sim...

Yes, best wishes to all of you.

Try engineering profiles? This could be very interesting for testing and demonstration toys. But I wouldn't do it for functional parts in equipment. The balls are very likely to go sliding instead of rolling, wear out soon, and even melt if printed in PLA.

I guess that could be a cooling issue. When printing small items, the hot nozzle stays on top of the model, keeps radiating heat, and prevents the model from solidifying well. Causing deformation and overextrusion. This also happens when switching from a huge layer area to a tiny one, and the printer slows down to its "minimum layer time", for example when printing a thin exhaust pipe on top of a huge building. Try printing much cooler, slow and in thin layers, so each layer has more time to cool. Or print two models next to each other, so the nozzle is moved away half of the time, allowing the model to cool down. Or print a dummy tower next to the model. I am not totally sure this is the cause, but it is worth checking. Printing a dummy next to the cones, reduces the overheating effect, but does not eliminate it: Theoretical concept of a dummy tower to allow cooling, and keep printing time per layer as constant as possible: A dummy cube in a real model:

Wear also depends a lot on what materials you print. Filled materials (metal fill, sand/stone, carbon fiber, glow in the dark,...) tend to wear out brass nozzles very quickly. Even plain white wears out nozzles, more than other colors, I guess because some are filled with chalk to make it opaque white.

If random, it looks like a mechanical thing, or unstable electronics. With power off, check if the head moves smoothly and easily, without hard points? Too much friction could also be a cause. For the belts and pulleys: with a marker pen, write a line from pulley to rod. So you can see if they move. On an UM2 with SD-card, I would say: check the card and slot for dirt. But I guess the S5 has an USB-slot, that shouldn't cause issues. If via network, maybe unstable connections and data loss? In the old days, when a driver overheated, it could for a moment switch off and cool down. But I haven't heard about that in the last years. If there are cooling fans for the electronics, check if they are working fine? No smell of hot electronics?