geert_2

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?

Thanks for the very informative videos. Also, the clock is a good idea. I guess the reason why airbrushing did not work very well, might be that the product just evaporates as soon as it comes out of the nozzle, instead of being deposited on the model. To me, an airbrush looks very similar to an injector in a car engine, and the injectors are there to evaporate the fuel as quickly as possible, without depositing it on the cylinder walls... It is not easy to see on the videos, but it appears that the acetone works better in fine details, and the methylene chloride works better on big surfaces? If so, depending on the model, it might be worth doing both? But that might also increase the risk of later splitting due to the chemicals?

You can reduce the ugliness a little bit, but not eliminate it, by designing your own supports. Then you can get them closer to the underside of the design, and still remove them well. There have been a few topics about this recently. Search the forum for: custom supports

A microwave heats molecules by applying a fast changing electric field. Thus the molecules are violently shaken left and right, which causes friction and warms them up, like rubbing. But this only works for molecules which have a polarity, like water, H2O: both H-atoms (hydrogen, positive) are sitting on one side , and the O (oxygen, negative) is on the other side. So, an H2O molecule has a clear polarity, and can be shaken by an alternating electric field. Food also contains a lot of water, so that is why it works. But I don't know if PLA has some polarity? If yes, it might work. If not, it could only work as far as there is moisture trapped in the PLA. Is there anyone who knows more about the chemical composition of PLA? Anyway, heating in a microwave does not work at all for dry non-polar molecules, like glass and most oil-based plastics. You can take them out of the oven with your bare hands. If you want to be sure, a conventional oven (well controlled, like an incubator), or the radiator of your living room, or an incandescent spot, might be a better heat source, I think. Or heating "au bain marie": in a closed dry box or pot, with desiccant, which is swimming in warm water. But not directly in the water, since that would increase hydrolysis. Anyway, it are interesting experiments. I would really like to know the results.

Keep in mind that standard breathing masks (fine dust filters, usually white) do *not* eliminate any chemicals from the air. They only filter out dust particles and paint particles. Which is good of course. But toxic gasses go straight through. To filter out toxic gasses too, you need a real gass mask like those used in hazardous environments in chemical plants, plus an appropriate filter for the type of chemicals you are going to use, and for dust. Filters don't filter all: the active chemicals in the filter are designed to trap specific pollutants only. These filters need to be changed whenever you start using the mask in the morning (thus one per day), and after a specific amount of hours of use during that day. And they need to fit airtight around your head, so: no glasses, no beard; or a special gass mask accomodated for these. Google for: gass mask. The sort of masks that make you look like an alien. We had those when I worked in a chemical plant long ago.

If I understood it well, there are two factors that make PLA brittle. But I am no chemist, so my knowledge is limited. One factor is hydrolysis: water is absorbed into the PLA and destroys its molecules. I think it does so by breaking the PLA down into lactic acid first, and then furter into CO2 and H2O? This damage is irreversible. After all, PLA is bio-degradable. And it is used in surgical wires too, where this chemical breakdown is used to make the wires be absorbed by the human body. However, just to try, I made a PLA sift in the sink of my laboratory, thus sitting in dirty water, but this has survived quite well in a year. The second factor is crystal structure: over time, PLA changes into another crystal structure that is harder and more brittle. But this can be reversed by heating the PLA. So if the PLA filament became brittle due to chemical breakdown due to water-absorption, it will give weak and brittle prints. If it became brittle due to changes in crystal structure, this should be reversed and should at first give strong prints like fresh PLA. In real life, both effects will play together, but I have no idea which is strongest. But at least you can minimise the hydrolysis by storing filament dry. Additionally, if the filament is really moist, melting it in the nozzle may cause water vapour bubbles in the print, which can greatly reduce strength too. Very common in nylon and ABS. If the PLA filament was stored dry, and it became brittle due to crystal changes, warming it up for a day (at 50°C or so, not sure?) should make it flexible again. Should. At least it should become flexible enough to feed well and to print. However, after one or two years, crystal structure in the printed part will change again, making it harder and brittle again. And hydrolysis will also play a part in the printed model. If the PLA filament became brittle due to water absorption and chemical breakdown, heating it will not help. That is as far as I understood it. Anyway, in two years time my printed models did become much harder to the feel (far less flexible), and sometimes a bit dull (I guess this dullness is due to chemical break down?). In models that need flexibility such as clamps or snap-fits, this is a problem. For demo-models it is not. Just my experience. If you use PLA for RC-model airplanes, or for structures under load, these effects might also cause problems over time. Stored fresh filament also becomes harder and stiffer, but I haven't had any feeding problems yet. But I do store them dry, in a sealed box with desiccant, so this is probably mostly due to the changes in crystal structure.

I don't know if brittle PLA filament breaks at a lower or higher forces than fresh and more flexible PLA. But if you have circumstances where the model has to flex to function, then brittle will break faster. For example: a clamp that has to snap around a tube. Fresh PLA filament will flex enough to snap around the tube smoothly. Brittle filament will no longer flex, so you can't snap it around the tube, and you keep pushing, and then it breaks. This applies to all sorts of snap-fit locking systems. Idem for keychains or card holders in your pocket. If you go sitting, the model may get stuck in your pocket and undergo huge forces. If it is very flexible (like nylon), it will bend and survive. If brittle (like old PLA), it will break. Don't ask how I know this... But even "more flexible" materials like PET do break in the latter case. Thus the PET or PETG filament for 3D-printing is not like the PET from cola bottles, which do not break so easily. So you have to take the practical application in consideration too, not just the technical strenght numbers. I think, but I am not sure, that flexible filament is better able to distribute applied loads, while hard filament gives higher local stress concentrations. For example, a scratch or an indent (like in chocolate bars) will cause higher local stresses in stiff materials than in flexible materials, and thus it will fail easier on that spot. I guess. If there are material scientists here, feel free to comment and to educate us.

Yes, I have used colorFabb PLA/PHA in various colors: white, red, orange, yellow, natural, bluegrey,... It works very well for me. I don't know if there are any other brands that use PLA/PHA. Also Ultimaker PLA works very well, for example Pearl. But I don't think it contains PHA, probably only PLA plus some additives to make it less brittle than pure PLA.

Here you have it, seen from both sides. One is from the first printer, the other from the second. Basically, it is just plied steel wire (inox spring steel, as used in dental appliances), so that it clamps around the edge of the frame. It looks primitive (well, it *is* primitive), but it works fine for me since 1.5 years. At first I also grabbed the priming blob with tweezers. But sometimes I forgot or I was too late, when doing something else inbetween the warming up. The reason why this priming blob does not stick to the glass, is that the purging is done from some height, at least on my UM2 (non-plus), so it cools down too much and is not squeezed into the glass.

For an UM2 (non-plus) try this: This wipes the nozzle after priming, so it does not drag the priming blob around anymore. I plied it out of inox spring steel wire (1mm), after an idea from another user who made it first in plastic (search for "filcatch" or something). The advantage of metal is that it does not melt, obviously. The spring steel makes that the nozzle easily slides over it, without causing damage. It just gives a nice ping sound. It has been in use succesfully for about 1.5 years I think, I haven't had any problem since. I don't know if you can use the same idea on other models (UMO, UM2+, UM3), because they might have different priming positions or starting routines.

Hello all, I just put the following designs online: 1) A compact spool holder for the UM2 which takes much less space than the original. Can be handy for narrow cabinets or tables. It has a standard 608 bearing for frictionless movement. My first version (with 4 pins) had been in use succesfully for almost 2 years, this latest one (with 3 pins, handier) for a few months. 2) A clamp to prevent filament from unwinding while in storage, and from falling off the spool sideways while printing (the last function only for colorFabb spools). Has been in use for a few months. 3) A character set and several character testplates for adding small 3D-text to a model. Character height is 3.5mm, width is 2.0mm, and line-width is 0.5mm. The testplates can be very handy for conceptual tests, to see which shapes and which heights work best for your models, even if you would design your own fonts. I made all these designs for myself. But they may also be useful for others. So, feel free to have a look and try them out: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ The original design files are provided in the RSDOC-format, the native file-format of DesignSpark Mechanical (very good freeware). Also provided are JPG-files showing the design, and print-ready STL-files. See the txt-file inside each zip-file for more info and instructions. All designs are under a CC BY-SA license, so you can use them for whatever purpose, also commercial. You can also modify the designs to your own wishes. But if you redistribute your derivation, you need to do so under the same license; thus you are not allowed to restrict the free use.

I have no experience with metalfills, so this is all guessing. Copper being one of the best conductors, I could imagine that the heat travels up into the filament really fast, and it gets through the teflon piece designed to act as a heat barrier. So it might weaken the filament before it enters the nozzle, and thus making it expand due to the pressure of the feeder? Like when you try to push a rubber band in a keyhole? Try heating a piece of filament with a soldering iron, or with one end in hot water, to feel how the heat travels? Or remove the bowden tube at the head, and feed a piece of filament manually, at different temps and pressures? Maybe you can get a feel of what happens?

I clean the glass plate with warm tap water first, with a paper tissue. And then I put a few drops of salt water on the glass, and gently wipe and keep wiping until it dries. The salt should not be visible as such. It should only be a very thin almost invisible mist. Like a wine glass that has been sitting in a cabinet for a year without using: if you look through it to the light, it has some sort of thin grey mist to it. That is how the glass plate should look for best bonding. Also, it does not work equally well with all PLA: the ICE brand still worked, but occasionally corners did lift. Ultimaker and colorFabb gave an excellent bonding. But if the lifting always occurs in one corner, it could also be that your bed calibration could be improved in that corner? Or that the glass may not be totally flat: even if it is only 0.1mm unflat, that is half of a first layer of 0.2mm. For PLA (Ultimaker and colorFabb), I mostly use 210°C nozzle temp, 60°C bed temp, 50mm/s speed, thus the default values on my UM2. First layer 0.2mm thick, the rest 0.1mm. For higher accuracy, I use 195 to 200°C nozzle, 60°C bed, 25mm/s speed. I never have lifting like in your photo. If the bed temp is too low, the model gets stiffer, but bonding is less. Below 50°C bed temp, it may suddenly pop off. If the bed temp is too high, the plastic stays too soft, and the model may be peeled off, in addition to sagging and deformation. I have seen this at 70°C bed temp. For my sort of models, optimal bed temp is between 55°C and 65°C. But I think gr5's method (10% dilluted wood glue in water) and neotko's method (strong hair spray) should also work equally well for these models. These methods are only a bit more messy to remove the models (that is why I prefer my salt method). Both have shown bigger and more complex models that worked very well. So if it doesn't, I tend to believe there is still some other unresolved issue?

I only use it in our university laboratory, and only for a minutes or so, until the model is filled and the gypsum is distributed nicely. (Although, since 3D-printing, I haven't used it anymore.) This sort of noise is more acceptable in a lab. But in an appartment, people would notice it a couple of floors away, just like drilling into concrete walls. Unless it would be placed on thick antivibration mats. The gypsum first looks like dry clumps of cement, but when you switch on the shaker, it starts flowing smoothly, very similar to silicone for mould making or lava, into all corners of the mould.

What about just keeping the old forum in use, parallel to the new one being set online? Until the new one is stable? Seems like the easiest thing to me? Are you planning to convert all existing threads to the new forum? Or is the old forum going to become read-only, as a sort of archive of accumulated knowledge?

I did sand a tiny injection needle down to 0.39mm, after cutting off the sharp point. So it just fits into the 0.4mm nozzle. After doing an first atomic pull, I carefully insert that needle from below into the nozzle, to push out any remaining debris in that small opening. And then I do another atomic pull to remove loose debris. In this way the opening keeps its internal diameter of 0.4mm. Apart from that: I don't really do "atomic pulls" anymore, I rather do "atomic turns": instead of brutally pulling the filament out, I first let it cool down much more, until 25°C. And then I wiggle and try to rotate that piece of filament. Then I heat it up again to 80°C (for PLA), while I gently keep trying to rotate it. As soon as it starts moving, I gently pull. No brutal force, only gently wiggling, turning, pulling. This seems to clean equally well. And yes, I do get the desired nice cone shapes, including a little 0.4mm tip. Advantage: no risk of damaging the rods or dislocating the nozzle by using brutal force.

This is the gypsum shaker I use, two speed. Small but heavy. Occasionally, it is capable of bringing a whole wall into resonation, like a pneumatic concrete drill. There is no way to get a *thick* gypsum paste flowing into fine teeth details without a good shaker. That is why I think it might be beneficial here too, if you have flowing problems. For dental models to be strong enough, the gypsum needs to be a very thick paste like cement or choco (it feels like somewhat inbetween). It is way thicker than honey or syrup, but less stiff than most natural clays. So they can not use a "creamy" liquid gypsum: that is too weak and brittle when cured. Normally, the mixing volume of (dental) gypsum is about 30g water per 100g gypsum, although it differs from brand to brand. Also, keep in mind that gypsum at this ratio slightly expands when curing (ca. 0.2%). And it may get warm to hot. If mixed in the correct volume, it is dry after curing, since all water is chemically bonded. But for use as filler like here, it does not need to be so strong, so it can be made more liquid, thus containing excess water. To slow down curing, you could try using very cold water from the fridge. If the gypsum is very dry (desiccant!!!), you could put it in the freezer or fridge too. But don't do this if it is moist: if you get condensation, that might already start curing it in the fridge.

It is not a font. It is a design I made in DesignSpark Mechanical, thus flat surfaces that I can pull into 3D-shapes. Created on a 0.25mm grid. I don't mind other people using it, but you do need DesignSpark Mechanical, as it is in its native RSDOC-fileformat. I will put this character set online (probably begin next week), together with some other stuff. The way I use it, is by copying and moving required characters into the desired text, and snap them on a 0.25mm or 0.5mm grid for alignment. And then I pull this into 3D-text, and apply it to the model, by adding or subtracting as required. Obviously this method is not suitable for setting huge quantities of bodytext for a newspaper in 3D. But for a single copyright notice or an On/Off label, it works well enough.

I have no idea what IRC is... I can still work with the current forum, provided I take the route "Community > Top categories > and then chose some category". Not optimal, but usable.

What if you print a test block of 20mm x 10mm x 10mm, 100% filled? And then 20% filled? Is there any difference in surface quality? I have seen similar blobs and gaps in some PET materials, when they needed to close big gaps: the material was too rubbery, instead of creamy. And thus, instead of pulling a nice line like PLA, it contracted into a blob on the nozzle that gets deposited on the model as soon as it reaches the next wall or so. So a 100% filled model would go well, but a 20% filled model would come out ugly with gaps. Bridges also would not work well. I have no idea if this is the cause in your model, but at least it looks familiar.

It really needs to vibrate hard to get a thick gypsum paste to flow well. Like a small combustion motor, a one or two cylinder engine. Putting the models on top of a washing machine will not work.

Forgot to say in my previous post: raised text (lines on top of a surface) usually comes out better than recessed text (hollows). In recessed text, corners tend to be cut off due to the elasticity of the molten material while printing, and the printer always has trouble with characters like N, R, K, 9, etc... Especially the capital N: the 0.4mm nozzle can't get into the sharp corners (remember that the characters are only 3.5mm high), thus this gives a very deformed diagonal line. In my tests, printing text on side-walls gave far worse results than on the top, because all corners are rounded to 0.4mm (at best), thus losing all fine definition, and the layer lines and overhangs interfere with the characters. It becomes a blurr. Text on the bottom also does not work: due to the squeezing of the first layer, and due to slight elephant feet for good adhesion, it gets closed down. However, "watermarks", thus voids totally inside the material, do work well. At least if the material is transparant or translucent, so you can see it. Solid text surrounded by a void works slightly better than hollow characters. In this case I recommend to make the text between 0.5 and 1mm high, for characters of 3.5mm and line-widths of 0.5mm. See the images. Why lines of 0.5mm width instead of 0.4mm for a 0.4mm nozzle: the conversion from geometric shapes and splines to STL-triangles does cut corners, and then on some spots the line-width will get less than 0.4mm. This causes trouble in the slicer: it doesn't print lines smaller than 0.4mm (unless you start messing with other parameters). If you go for 0.5mm lines, you don't risk this, and it doesn't matter if the line ends up 0.48 or 0.52mm. Use fonts with equal line widths, or at least use bold or heavy fonts. I designed my own simple character set for 3D-printing: it took less time to draw this once, than repairing other characters in the 3D-design every time.

If I go via "Forum" I also have these problems. But what still works well is to go directly from the top menu "Community" to "Top Categories", and then "General", "The art of printing", etc... So I have to avoid the direct path to "Forum". Browsing the forum without loging-in, and then only log-in when I want to reply, also works fine, but is not handy.

Thanks for the video. Yes you are right: a piece of art needs sufficient weight to feel like art. Otherwise it feels like a cheap plastic toy, even if it looks identical. A few questions: why not use a 20ml or 50ml syringe? Then you can inject it in one shot. In Belgium such syringes are available from our local pharmacy (way cheaper than from companies that sell laboratory equipment). Instead of a needle, have you tried drilling a bigger hole, and using thin aquarium PVC tubing? Or no needle at all, just the syringe nozzle? Then you can make the mix thick like cream, and still have it extrude well. This will give a more sturdy and heavier filling, and it does not leak water. When making dental gypsum models, the laboratories put the model on a vibrating plate: these strong vibrations make even very thick gypsum flow like liquid lava into all corners. It also helps the trapped air bubbles to float to the top and spat open. So if you would have something around that vibrates hard such as a small combustion motor, or a cylinder compressor (mount a flat metal plate on top, so you don't spoil gypsum into the compressor), that might help. I have a dedicated dental vibration plate, but that was not cheap. Or maybe you could try an electric tooth brush without brush, or even a female "massage tool"? Anyway, keep posting videos and pictures.

I wipe the nozzle's outside with silicon oil and PTFE oil sometimes. In the beginning I had to do this more often, but now I do it occasionally. It reduces accumulation somewhat, but does not totally prevent it. It also depends on the material you use: PET caused more problems than PLA for me. Also, if you have overextrusion, this happens more, as the excess material accumulates on the nozzle. Printing cooler and slower also helps a bit for me. (This is for an UM2 with 0.4mm nozzle.)