[Smooth Top Surface - Best Case Expectations]

Depending on the situation, another good option to minimise these lines, might be postprocessing and chemical smoothing. See the thread on "Acetone smoothing" by user Cloakfiend who has done the most experimenting on this, and who achieves excellent results routinely. His models have virtually no visible lines. It's a lot of reading, but then you get a wealth of info too. 

 

[What can i do about that ?]

First make sure that the bed is leveled correctly and is very clean (clean it with pure water, no soap, no detergents, since soap reduces bonding). Then you could try this: wipe the glass plate with a tissue moistened with salt water. Keep wiping gently while it dries. This improves bonding, but only works for PLA.

 

For the full manual, see: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

 

For other materials than PLA, you might need to use glue (10% while wood glue dilluted in water, user gr5's method), or hairspray (user neotko's method).

 

[Underextrusion with Extrudr Green-TEC]

I don't know Green-Tec, so I can't comment on that.

 

But have you tried to remove the bowden tube at the nozzle, and then feed the filament manually, while adjusting the temp in steps of 5°C? This might give you a better feel of how it behaves, and at which point it starts to flow well.

 

Of course, a few atomic pulls might also help. It could be that the nozzle isn't blocked, but the opening is a bit reduced, or that the internal walls are covered by burned residu, which hinders heat transfer? You could try my more gently atomic pull method: remove bowden tube, manually heat the nozzle to the normal printing temp, push through a bit of filament, then dial down temp to zero, let filament cool down to room temp, and then gently wiggle and turn the filament. No brutal pulling. Then heat the nozzle up again to 70°C (for standard PLA, other materials require different temps, use trial and error), and while it is heating up, gently turn the filament until it comes out. Always gently, no brutal force, so you don't risk bending the rods or displacing the nozzle parts. The long manual with pictures is here (scroll a bit down): https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

[Problem with material flow UM2]

4 hours ago, Massassi said:

How often should I change the teflon, (how many print hours). are there other parts that need replacing?

When it is worn out. That depends on the temperatures at which you print. Once above 210...220°C, the old teflon couplers degrade quite fast, especially when printing fast (=high internal pressure). When printing slow (=low internal pressure on the molten filament) and cool (=less thermal deformation and break down), then it lasts much longer. I usually need to replace them every 200 to 300 hours on my UM2 (non-plus), when mostly printing PLA at 210°C, and PET at 220°C. Apart from that I haven't replaced anything yet, and both printers have about 1000 hours. Although I made a few minor modifications, such as a compacter and low friction spool holder, and different bowden-tube clamps (to make them pinch the tube less).

 

[Removing support structure from small, deep holes]

If the hole is only 3 to 4mm diameter, you usually do not need supports for most materials. If the design alows it, you could taper the holes near the end (like in blind holes drilled), to improve the inner shape, if that would be required.

 

For a bit larger holes, let's say 10...15mm, I design custom supports that are strong enough to wiggle loose and to pull out, and that have enough clearance to insert tools next to it, or that have holes where I can insert hooks to pull, or something similar; it depends on the design.

[Hole through gear]

If you are familiar with SketchUp, then another free editor you might want to try is DesignSpark Mechanical. It only requires registration, which is a fair price. It is a limited version of the commercial Spaceclaim. The user-interface is very similar to SketchUp, and there do exist lots of good training videos. (And you can also use Spaceclaim's, if you keep in mind the limitations.) I have made hundreds of models, and *never* had any problems with the STL-files.

 

[Disappointing second print with UM3E, stringing]

I think for 3D-printing it might be possible to reduce file size to a few 100MB, and still have good quality, as very fine details are lost anyway in a 3D-print?

 

[Does this "math art" vase require support and/or a bed?]

Indeed, concerning "base" I meant something like a brim. Or a custom made brim-like plate, maybe with a few custom made tree-like support columns to support the overhanging middle parts, if you print it in one piece. If you print it in two pieces, cut halfway, this might not be necessary.

[Disappointing second print with UM3E, stringing]

That is a really nice model. Where did you get that?

 

An alternative to printing the globe at once, with supports, might be to split the model along the equator, print both halves separately, and glue them together? But then you will need to post-process the seams. For good stability, I would make the shell thick enough, maybe 1.2mm?

[Material not feeding through]

What I would do in that case is:

 

- First do a couple of "atomic pulls" to clean het nozzle. Then you are sure there is nothing blocking the nozzle, and the inside of the nozzle is not totally covered in ashes which might hinder heat transfer to the filament. Also, by the shape of the molten part, you can also see if your teflon coupler is worn out: this will cause a sort of blob or lip deformation in the molten and cooled filament, near the edge between the brass and teflon part.

 

- If this blob or lip is present in the pulled out filament during atomic pulls, you definitely have to replace the teflon coupler. Using the old coupler, I had to do it about once a year, for printing mostly PLA. As said above, use the new version of the coupler.

 

- If the PLA is old, it will get stiffer, and more difficult to unwind from the spool: it acts like a very strong spring that resists unwinding. And it causes a lot of friction in the bowden tube and in the nozzle. These all contribute to underextrusion. A simple solution is to manually unwind a bit of filament, bend it around a skater wheel (7cm diameter) in the *opposite* direction, so it straightens out. Then unwind and straighten the next 15cm, etc., until you have done a few meters. This only takes a few minutes, and I do this while the printer is warming up. Then wind it up on the spool again loosely, so that the bending radius now is about 30 to 50 cm, instead of 10cm before. For the old UM2 printers (non-plus version), this helps a lot.

 

- For a more gentle atomic method to clean the nozzle, see my manual at: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

This does not require brutal pulling, but only gentle wiggling and rotating, and deeper cooling.

 

- With compressed air, clean the feeder wheel from dust and debris, so you get good traction.

 

These things will most likely get you going again, I think.

 

[Cadkey solids]

Cura can not directly open 3d-design files.

 

You need to export your design from your 3d-editor to the STL-fileformat. And then open that STL-file in Cura, or in any other slicer of your choice. And then save the result as a gcode-file for the 3d-printer.

 

For exporting to STL, usually a few standard quality settings will be available, depending on the editor (I don't know Cadkey). Try medium quality for big objects with not too much details, or fine for objects with much detail. Of something equivalent. This is trial and error to see what works best for your typical designs and your 3d-software.

 

(But of course, before exporting, first save your design in your editor's native fileformat, so you don't lose it.)

 

[Tolerancing for fitting parts?]

Most of the time I make a gap between parts of 0.1mm for a tight fit (requiring force to insert), and 0.2mm for a normal or loose fit. This is usually sufficient for the normal irregularities due to layers, rounded edges, ringing effects, occasional small blobs, slowing down on corners, etc... For small gaps, I take larger tolerances, based on experience, or after trial and error. For example, a 4 mm hole usually ends up as a 3.5mm hole. Then the best solution is to go through it with a 4.0 mm drill (manually!!! Never electrically, this melts PLA). Rounding bottom edges with a 0.5mm round, reduces elephant feet.

 

Printing slow and cool improves tolerances. Printing hot and fast makes it worse, since the pressure can not change immediately in the nozzle upon stops, retractions, etc. So that causes more irregularities in the print.

 

[Soft PLA troubles]

What you could also do to diagnose the problem, is to remove the bowden tube at the feeder end, and manually insert and move filament. Manually dial nozzle temp up to various settings, and extrude some material. Then you might be able to feel where it gets stuck and why?

 

[Offset Print Problem]

It could also be that the X and Y stepper motors did miss a few steps, when they hit something. They have no feedback, so they don't know where they are, they just count steps. And thus if they miss a few steps, if they can't move due to too much friction, all the rest of the model will be shifted.

[Ultimaker 4, The next generation of 3d printer]

An estimation of the required power could be: power consumption in Watt of your equipment (UM3 + any other equipment that has to stay on) x hours to survive power outage x 3. Thus: Watt x hours x 3. Then this would account for small variations, for degenerating batteries over time, and for all sorts of technical details.

But if you only have one power failure in 10 years, it might be cheaper to simply redo the print. UPS batteries don't live forever: expect them to fail in 5 years, probably at the time you need it...

And if you calculated the capacity to survive let's say 2 hours, but the power failure lasts 2.5 hours, then your print still fails.

Usually, the main purpose of an UPS is to survive *short* power failures, so that you have time to safely and correctly shut down your equipment, without loss of data, and without file system damage. And to smooth out spikes or irregularities in the mains voltage.

So, personally I wouldn't do it.

[Bad walls (gaps) when layer time changes (UM3E)]

This is caused by drastic changes in layer cooling time, due to changes in layer-area to print. You will notice this especially on small objects, or on objects where the area to print suddenly changes from very large to very small. Exactly like the objects you show.

A solution is to place a dummy cube next to the model, and hollow it out at appropriate heights, so that the total cooling time per layer is almost identical for each layer. The simplest method is to put a copy of your model into the dummy cube, and subtract it. However, make sure the dummy has a bottom surface layer covering its total area, so it sticks well to the glass plate, otherwise it may be knocked over. Or manually design the dummy.

Printing slow and cool also helps minimising this effect, but this alone can not prevent it.

I often need to use this method, due to the layer area of my models suddenly changing from ca. 10cm^2 (=1000mm^2) to only a few mm^2.

[Print Text Without Base Material]

I have never tried this approach via images, so I can't comment on that.

But if you want to print 3D-shapes, letters or whatever else, I would suggest you design them in a 3D-editor. Then you can get everything right in the 3D-editor itself. And you can slant the edges, etc.

I use DesignSpark Mechanical. This is good freeware by RS Components, and only requires registration. It is a limited version of the commercial SpaceClaim 3D-editor. This has an easy to learn interface, similar to SketchUp. There are lots of good training videos available on Youtube. In a few hours you can design simple objects. In a few days, you can design complex objects.

However, it has no built-in text features, so you have to work around: or draw the characters yourself, or import them from another editor, or use the "dimensioning tool" instead. Google for: how to make text in designspark mechanical.

Try to avoid SketchUp: this causes problems, since a lot of shapes are not "watertight": thus they are no solids, but sort of "cardboard" models with infinite thin walls.

You can set the text in SketchUp, export it, and import the skp-file in DesignSpark Mechanical. But then you have to repair these defects due to non-closed vectors manually first. And then in DesignSpark Mechanical, you can extrude it into 3D, slant edges, and do whatever else you want.

Before trying any editor, watch a few demo-videos on Youtube first, to see if you like the workflow.

[Please develop an archival PLA surface finish that prevents moisture resorbtion]

Do you spray it with an airbrush, or do you use spray cans? In both cases, could it be that the paint is dilluted with solvents, and that these solvents cause the problem? A bit similar to the "acetone smoothing" that cloakfiend uses for his models, which also causes whitening, but then in this case as an undesired side-effect?

If so, manually brushing on a varnish might improve things? But these still contain solvents, I think...

Anyway, for outdoor use you would definitely need an UV-absorbing and resistant paint or cover.

Printing the model in 3D, or printing a mould, and then casting in plaster, cement (for statues), or UV-resistent PU (for toys and tools), might also be an option?

Edit: forgot to say: the most UV- and weather resistant varnishes and paints available here in Europe, are yacht and ship paints and varnishes. But I have no idea if they work on PLA.

[Does this "math art" vase require support and/or a bed?]

With that amount of fine details and holes, I would print it slow and as cool as possible. This will also reduce curling up of edges. Maybe 190°C at 25mm/s, if your PLA can handle that? Maybe you could cut out a small part and try that first, and then manually adjust temp, to find the best settings? Also, be sure to have a big thick stable base, otherwise it will be knocked over.

[Problems with warping by shrinkin holes]

As gr5 said, I also found that printing in the left-back corner is more prone to problems. I am not sure , but I think this might be due to the leveling, to light warping of the glass (even if only 0.05mm), and to the tight bending radius of the bowden tube in that area: this causes much higher friction of the filament in the tube and nozzle, which might lead to slight underextrusion, and thus less good bonding?

Anyway, I get best results when printing in the center and front-right areas. This is for my UM2 (non-plus) printers, which are more sensitive to feeding problems than UM2+ and UM3.

Also, you could put your models closer together by alternating them, so that the wide parts of one model slide into the openings of the other model. If your Cura-version would not allow this (I have no experience with the latest versions), you could do this in a 3D-editor. For example, if I would have to print two "E" characters, I would turn one upside-down and shove that tightly into the other, so it occupies only half the space on the build plate.

Concerning optimal bed temperature, I did several tests some years ago, and I found that a too low temp makes the models pop-off suddenly. And a too high temp makes the models too flexible, and makes its corners peel off the glass due to warping. You have to find the exact spot inbetween, where the bonding is still good, and the model stays stiff enough not to peel off. This is different for each material, and might also differ a bit from printer to printer. You could place a test object in the most critical area, and print that with changes in bed temp in steps of 5°C up and down, until you get both effects. Stay around to see what happens. Then choose a value halfway inbetween. For my Ultimaker PLA and colorFabb PLA/PHA, 50°C is too low, and 70°C is too high, for use with my salt method (may be different for other bonding methods!). I need to stay between 55°C and 65°C, thus I just leave it on 60°C, the default.

[Massive prints (More than a kilo)]

As JohnFox already mentioned above, my first idea also was: print only a shell, and fill that with resin? Polyurethane or so? There do exist lots of PU-variations: from hard and brittle, to soft and flexible. Be sure to use a resin that cures slowly, with low shrink, and that does not get too hot, otherwise the shell might melt, or the whole thing might catch fire and/or explode.

Or print the shell with very light infill, just enough to make it stable. Post process it (sanding and smoothing), make a silicone mould, including a hard shell around the silicone for stability, and fill that silicone mould with resin. Be sure to spray the silicone with release spray: this gives a much longer mould life. If you need to have more than one copy, casting could be the fastest option. You can reuse the mould, or you can even cast multiple moulds from the same 3D-printed part.

On Youtube you can find lots of good videos on silicone mould making and casting.

[Does brittle (easily snapped) PLA filament equate to less strong prints?]

 

Won't this type of cracking be eliminated in the "furnace" of the hotend?

 

That is a good question, but it may be that the material itself does not actually regain its original composition/qualities once it has aged.

I do not know enough about material sciences to answer that. It would be nice if one of the Team Ultimaker guys could give us some feedback on this.

 

When this material is molten, the cracks disappear and it flows smoothly out of the nozzle, like always. So, in my prints there are absolutely no cracks indeed.

But as kman says, I also wonder how much strength or flexibility old filament recovers, compared to brand new material?

I guess if the material became harder and brittle due to reversible changes in crystal structure only, it might recover fully. But if the material became harder due to hydrolysis (break down of molecules), or by plasticizers that evaporate or degenerate, to me it seems unlikely that it would fully recover. But this is only a guess. I would welcome the views of a materials engineer specialised in PLA polymers.

Anyway, printed parts that get old, do get much harder and brittle, in my experience. This is very clear in my clamps. For example the "snake clamp":

When freshly printed, this easily flexes and smoothly slides over the bowden tube. But after 1.5 years now, they are so stiff that they are very difficult to get around the tube. And sometimes they just break. I have seen this too in other designs like keychains, which I made specially to test real life behaviour. I have no scientific data and no numbers, but the subjective feel is obvious. So this is something I keep in mind when printing parts that need to keep their functional flexibility over a longer time.

I will see if I can find a few old unused clamps, and warm them up in the oven (=incubator, electronically controlled) for some time at different temps, and see what that does.

[my UM3 Extended is experiencing multiple errors after being moved to a new location]

Have you checked if all connectors are still well inserted? Just every connector you can find? Gently (!) wiggle it and push it back on. Also visually check if everything is still aligned well and no corners are damaged from rough handling.

After moving mainframe computers in the 1980s (they had many 21" racks then, see the photo), it was standard practice for us to open the cabinet, and to reseat all print boards, and all flatcables, from the very first to the last. Because they would shake loose during transport on a truck. This solved 90% of problems caused by moving.

[Does brittle (easily snapped) PLA filament equate to less strong prints?]

...

Something about straightening filament and holding it in that position for 10 hours makes it very brittle.

This is most likely caused by microcracks slowly growing in the material, due to the stress. See these pictures: first one taken with macro-function in daylight, second one taken through a 10x microscope under warm white LED-light. This is colorFabb "natural": due to its translucency the cracks are very visible, but I also see them in other materials. This is fairly old filament, about 1.5 years, although it has been stored dry in a sealed box with desiccant. These pictures were taken after manually straightening this tightly wound filament near the end of the spool, and then relaxing it again (so the cracks stop growing), to make it feed better in my UM2, without underextruding.

But I guess it is the same effect that plays when the filament sits "artificially straightened" in a bowden tube (or wherever else) for a prolonged time. I have been wondering too about the reason why it cracks? Is it spooled while still warm, and then cools? Or spooled cold, but the crystal structure takes some time to change from amorphe (more flexible) to crystaline (harder and more brittle), so it is wound while still in the more flexible phase, and then it hardens while already wound, and gets fixed in that bent shape?

These bright lines are cracks: you can see them beginning and growing while straightening tightly wound filament.

[Acetone Finishing on PLA]

A thought that just occured to me: most plastics do contain additives. For example: plasticizers, pigments, UV-stabilisers, flame-retardants, water-repellants,... I don't know if PLA or PLA/PHA contains any such additives (apart from pigments of course), but it is not unlikely.

Could it be that each of your solvents dissolves different parts of this mix?

For example, imagine that a plastic blend would consist of a mix of 50% hard plastic, and 50% plasticizer: if one solvent would remove the plasticizer, then the result would be brittle and hard plastic. If another solvent would remove the hard molecules, then the result would be soft and flexible plastic.

Maybe this might be worth trying? For example, if you print 4 small test plates (let's say: 50mm x 10mm x 0.5mm: this prints in a few minutes) and then thoroughly brush each with one of your solvents. Keep one untreated. And then see if there is a difference in flexibility and brittleness, now immediately, and in a few weeks/months?