RepRap 3D printers, PLA, and moisture problems

I’m not talking about making waterproof, foodsafe or dishwasher safe objects here (that’s a topic for later), I’m wondering about storage conditions for filament before I print with it. I’ve read a number of statements that suggest ambient humidity can be a problem.

PLA is hygroscopic. It absorbs moisture from the air. Here’s a patent application that discusses it as a problem for plastics engineering.  When PLA with moisture in it runs through the hot end nozzle of a RepRap printer at nearly twice the boiling point of water, the moisture turns to tiny little bursts of steam. Its a bit like shaking a bottle of carbonated soft-drink, then trying to pour it into a thimble. Not exactly precise.

The PLA filament I use for my printer is called 4043D PLA. The pdf  at Ultimachine.com on 4043D PLA in film (not filament) form states

“In-line drying is required. A moisture content of less than 0.025% (250ppm) is recommended to prevent viscosity degradation. Typical drying conditions are 4 hours at 175 degrees F (80 degrees C) or to a dew point of –30 degrees F (-35 degrees C), with an airflow rate greater than 0.5 cfm/lb of resin throughput. The resin should not be exposed to atmospheric conditions after drying. Keep the package sealed until ready to use and promptly reseal any unused material.”

Now that sounds pretty heavy duty, and is aimed at factory production. I’m guessing ‘viscosity degradation’ means ‘gets sticky and clogs’, but I could be wrong. More directly relevant to me is the RepRap wiki page on PLA, which I mentioned above. It states

“PLA can absorb moisture from the air. When it is heated this moisture can turn to steam bubbles which with certain hot end (extruder head) designs can interfere with printing. The symptom is that when the extruder motor stops the PLA kept coming out. When the stepper starts again there is a significant delay. Occasionally the tip may blow a bubble with a tiny puff of what looked like steam.

Small amounts of PLA filament (Natureworks PLA4043D has been tried) can have some moisture removed by putting it on a piece of aluminum foil in an oven heated to 170F for an hour. The filament in the oven is floppy, but sticks to itself only slightly. Flexing the coils after cooling unsticks them from each other. Heating a whole spool this way has not been tried, and may result in the spool becoming unusable, so caution is advised.”

“Sticks to itself only slightly” does not sound like a success to me, given how important consistent feed rate is for reliable smooth objects.

I don’t know how common this problem is, but I’m worried anyway. I live in a humid subtropical area where in summer the thunderstorms can drop a foot or so of rain over a couple of days, then the sun comes out and temperatures reach 37 degrees Celsius (98 degrees F?). Humidity >95%. We have no air conditioning. Mould on the ceiling is a problem if we don’t keep the doors and windows open for airflow. I understand that most people in developed countries tend to have air-conditioning, and I wonder if that, more than extruder head design, is what has been protecting most folks.

At our last remaining real photographic store in town I saw a small bar-fridge sized glass-fronted dehumidifying cabinet. I know photographers locally who have big problems keeping their lenses (some of which cost more than my printer) free of mould. Especially the two guys who live in the rainforest… That cabinet costs $400. Plus electricity to run it. I’m wondering what DIY options might work. A sealed fishtank with a small incandescent bulb and some desiccant? Am I just being paranoid, trying to think of things I can do before my MakerGear printer kit arrives?

All suggestions gratefully received.

Advertisements
This entry was posted in 3d printing. Bookmark the permalink.

23 Responses to RepRap 3D printers, PLA, and moisture problems

  1. I haven’t had any troubles yet, but then I haven’t carried my PLA through a summer here yet either.

    I think if it’s a problem, I’ll get one of those little 12v can coolers, gut it for its peltier and heatsinks, and make a dehumidifier with that. Basically the idea is to install it in a U-shaped piece of pipe with a fan, so air is first drawn over the cold side which condenses water out, then gets drawn over the hot side to warm it back up and return it to the chamber. The U-bend will have a bleeder valve at the bottom to let the water out.

    The peltier power must be carefully managed to keep the temperature below the PLA’s glass transition (60 celsius), so maybe running it for 30 seconds out of every few minutes or something would do the trick.

    I’m not sure how effective that’ll be, but it should certainly be better than nothing!

    Maybe leave the filament in there for a few days, then chuck in some rice which is a cheap and plentiful dessicator 🙂

    • BrazenArtifice says:

      Interesting idea. And nicely weird-science, what with cooling sections, and heating sections, and drippy-oozy sections, and temperature readouts…

      I see what you mean about careful power management. 60 celsius is not very hot. With the sort of variations in ambient temperature round here I’d want some serious watchdog timer or failsafe to make sure it can’t just melt the whole lot! Maybe it’d be a good use for one of the old BASIC Stamps I’ve got in a box somewhere.

      I’ve never worked with Peltiers. Vague memories from the seventies when I first heard about them suggest they are outrageously current hungry. Hopefully 30 years of development have eased that problem? While I do have mains power here (most of my neighbours are on solar), I wouldn’t want another fridge sized addition to the power bill.

      Is this idea your own, or is it worth my while googling for people who’ve built one and might have solved some of the design problems already?

      We had >40mm of rain in 4 hours on Sunday night, and now its >32 celsius. Just nature reminding me to do something about my PLA (which is still sealed up in its airtight bags with desiccant provided by Makergear)…

      • It’s my idea, but I’ve since heard that quite a number of commercial dehumidifiers work this way

        If you’re into weird looking science, make yourself a batch of cupric chloride for etching PCBs, it’s a most excellent mad scientist green! And when you’re regenerating it, you have to bubble air through it and it’s ready when it turns from opaque browny green to translucent forest green 🙂

        • BrazenArtifice says:

          I was pretty highly strung at school, so my hands shook too much to enjoy practical chemistry. Acids and electronics are not my favorite mixture, no matter how cool the colours 🙂 Veroboard forever! Fond memories have just popped into my head of an elderly all-valve HF transciever I bought during my Amateur Radio phase in high school. Getting my antenna loading wrong generated very mad-science arcing across the air-variable capacitors in the final amplifier. 2 inches of blue RF lightning on my desk!

          • I never did chemistry in school, I’ve basically built it on top of my physics understanding just like my violin playing. An etchant that eats copper and makes more etchant from it is very cool, thought I’d share.

            My reprap electronics are on veroboard, see https://picasaweb.google.com/lh/photo/oxSHmuY2Y84gtsdi2Tp5dg and https://picasaweb.google.com/lh/photo/aL_pHVdGiNe2rhwacUjNQQ . Looks messy, but works really well!

            As for 5cm sparks, I was making those with a car ignition transformer in primary school.. they work great up to about 100hz! They sure look different to RF ones though- more blue, less pink. I made a few tv EHT transformer flybacks too, they ran up at around 13khz or so, lovely pink-cored sparks albeit not as long as the car transformer.. used to connect them to lightbulbs to excite the argon and get the whole plasma ball effect going 🙂

            • BrazenArtifice says:

              Hey, Sorry that last post got sent to moderation hell. I’ve changed the wordpress defaults so it won’t happen again. Paranoid wordpress thinks only spammers put more than one link in a comment 😦

              Your violin playing is built on top of your physics understanding?! Just getting past the squawking phase on bowed instruments takes a huge leap of faith!

              And yes, thanks for sharing the cupric chloride info, I’d not read about it before. An eco-friendly re-usable pc board etchant is brilliant enough, that it changes colour with bubbling gas is, as you say, very cool. And I bow to your greater knowledge of arc-making. With me the shower of sparks was always a mistake (mains 240V is not a substitute for a 50Hz audio tone generator… oops.)

              As for the veroboard, it was my second ever form of scratch-building construction, after I realised that my bird-nesting on top of tag strips was insanely messy and horrible to troubleshoot. I spent some time looking for electronic CAD software that would design veroboard layouts for me, but I was never happy with the results, and always went back to working with blue-lined grid paper and a pencil. I’ve never etched my own pc boards. That acid phobia again…

            • Yes, my musical ability is entirely built on my understanding of physics, and a few fundamental understandings about modern musical theory. Basically, the violin is a tuned resonator that takes wide-spectrum noise from drawing something sticky over something taut, resonates at a particular frequency and encourages the noise source to produce more of that frequency. The body shape allows multiple resonance modes with standing waves at various orientations within it, and the frequency is selected by the tension and length of the string. There’s an optimum range of pressure and speed for the bow on the string, which changes with intonation and frequency, and also between strings. Finding this is empirical, and part of the learning process 🙂

              As for the musical theory, frequencies are perceived by us as logarithmic. An octave is an exact doubling or halving of frequency. Halving the length of a string will double the frequency, and halving it again will go another octave up. This is why the notes get closer together, the further down the neck you go on violin, guitar, etc. Combining any note with the same note in a different octave gives an exact power of two multiple which combines perfectly. Other harmonies are possible too, all the ones with small integer fractions (such as 3/5 and 2/3) sound sweet, whereas ones with larger integer fractions (eg 11/12) sound discordant. Irrational number ratios and enormous fractions sound positively dreadful, and they cluster thickly around the sweet ones! Finding the proper ones is again part of the learning process, and should be programmed into the various ganglia governing the fingers and arms.

              Our modern musical score divides each octave into 12 notes, equally spaced on the logarithmic scale- A, A#(Bb), B, C, C#(Db), D, D#(Eb), E, F, F#(Gb), G, G#(Ab). This doesn’t allow all chords and harmonies to be reproduced perfectly, but it gets us fairly close for the common ones. We’d need 53 notes per octave to get really close to /all/ the good ones!

              When playing the violin, the player must use their ear, knowledge of musical theory, and experience with the scale and piece of music in front of them to shift these notes closer to the appropriate perfect harmonies. The violin is one of the few instruments that can do this, and this is the reason that you should /never/ learn to play one using a chromatic tuner! It will tell you that you’re off, when in fact you’re on the perfect frequency for that piece of music!

              Anyway, I use eagle PCB editor for laying out my veroboard. I set the grid to 2.54mm and tell the autorouter to prefer horizontal tracks and that change of direction and using top layer is expensive. After finding a component layout I like using the autorouter by moving things to minimise excursions to the top side, I rip up most of its crazy madness and do it again by hand.

              Yes that’s PLA from vik of http://diamondage.co.nz , the closest thing Australia has to a local supplier. I’m running pretty low on the translucent purple and still have kits to print. I have some more plastic coming in, but none of it is that particular purple. I find that it’s more difficult than other colours to get the quality I’m looking for in my prints. My favourite so far is his deep purple, which you can see as the two pulleys in that pic. It comes out looking fantastic!

              I am planning on reselling vik’s filament here to save us all some postage, just haven’t hammered out the details yet.

            • BrazenArtifice says:

              Arabic and Classical Indian music are heavily microtonal, so I have heard. My understanding of the physics of music hasn’t helped my playing nearly as much as sitting next to real musicians in irish music sessions. I have guitars, button accordions, harmonicas, a mandolin, 2 elderly keyboards and a bunch of tin whistles in different keys, and I still can’t play any of them welll enough to play in public. Lack of dedication == easily distracted.

              I’m hoping to get to grips with KiCAD for circuit design, but I’m trying to avoid diving into a whole new lot of unknowns when I should be building my printer. Once it can make me weird organic-shaped boxes with animal-headed knobs for the pots, I want to play with making Lunetta-inspired CMOS sound synthesisers. Gah, my list of neat things to design and print is three pages long already 🙂

              I’ll definitely be up for some translucent purple PLA when you get the importing sorted. And Bronze, so I can make Brazen Artifacts! I got MakerGear to put three pounds of PLA in with my printer kit, and it only increased the postage cost by $10, since the box was already so heavy.

            • I fiddle while my printer prints, if I’ve nothing better to do. I’m certain the sounds that the motors make while printing will eventually influence my music!

              If you want to make things from bronze, PLA can be used instead of wax for lost wax casting. Got a kiln and forge up there?

              I’ve been meaning to learn KiCAD too.. I suppose I haven’t worked up the motivation to tackle the learning curve since I know eagle so well already

            • BrazenArtifice says:

              Lost wax casting will be fun, eventually. That’s one reason I made sure to buy some ‘natural’ PLA, as I hear it burns out more completely.

              Bronze is lovely, but classical age bronze had arsenic added to the copper to harden it. The fumes led to arsenicosis, and all the Bronze Age smith gods were ugly and lame from arsenicosis’ characteristic symptoms. Hephaestus even built a bronze wheelchair for himself. Modern bronze has zinc added to the copper, leading to metal fume fever, so is best cast outside. (Amazing what a head stuffed full of odd historical facts can lead to, when assisted by search engines.)

              Might try casting in the white metal that people use for casting wargames miniatures, though bronze polishes up so much better.

            • BrazenArtifice says:

              Just found your shop.

              The translucent purple mendel parts kit looks wicked. I want to build a translucent purple spaceship for my shiny purple otter stuffed toy (from “Bear in the Big Blue House”).

              Is that PLA? Where did you buy it from?

    • Daan Snijders says:

      What about putting the PLA in a microwave?
      I did not try this yet, but in theory it should remove moisture, right?

      • Microwaves cook food by heating up the water content, right? How can you be sure they wouldn’t heat up any absorbed moisture to above the softening temperature of the PLA? Still, I guess it’d be easy enough to test – just stuff some slightly damp PLA into a microwave, with a weight on top so you can see if it deforms.

        To be honest, I’m not having any of the problems I expected with moisture absorbance of my PLA, despite living in the humid sub-tropics.

        • Daan Snijders says:

          I did a test, and the moisture does heat the PLA above the softening temperature.
          A few meters of 1.75 mm PLA, in the microwave at 600W for 2 min.
          So this is not an option. Worth trying though….

  2. Jack says:

    I might suggest keeping PLA in a sealed plastic container with some desiccant like ‘Damp Gone’. Once it is consumed, it can be rejuvenation, but depending on your situation, just buying new may be a better/easier option.

    • BrazenArtifice says:

      Yes, that’d be my solution for long-term storage. With the experience of having just lived through our summer wet season, and not having any problems with moisture absorbing into any of my PLA, I think the best solution is this:-

      Print lots of things! (ie, use up the filament so fast it never has a chance to get wet).

  3. Hello,

    We have recently posted a Kickstarter campaign to help with this exact issue! Please review the link below for The Filament Safe, an airtight 3D printing material storage solution.

    Thank you,

    Stephen Wierenga

    Perception Engineering

    • Lyn says:

      My daughter works in the media center of a university. Their solution is to use a common food dehydrator to restore older video and audio tapes.

      http://www.amazon.com/Nesco-FD-61WHC-Snackmaster-Dehydrator-All-In-One/dp/B0002WSQHU/ref=sr_1_4?s=home-garden&ie=UTF8&qid=1405614191&sr=1-4&keywords=food+dehydrator

      I already had a 4 layer food dehydrator which I cut out the grating between 3 of the layers to accommodate reels of filament. The big advantage of a food dehydrator is the temperature range does not get hot enough to damage filament and the air flow is high enough to dry a reel in about 4 hours at 135F/58C. Maybe more or less time depending on the amount of moisture saturation.
      I purchased an “AcuRite 00613A1 Indoor Humidity Monitor” that shows current Humidity and temperature as well as the history of highs and lows from Amazon.com for $9.98 Prime. http://www.amazon.com/gp/product/B0013BKDO8/ref=oh_details_o00_s00_i00?ie=UTF8&psc=1

      I checked my new desiccant pads and they showed 16% humidity when received. I put them in the dehydrator and will store them in a sealed bag when dehydrated. After I dehydrate a reel I’ll store the reel in a double sealed plastic bag with a couple of desiccant pads until I use it, then I immediately put it back in a double sealed bag with desiccant pads.

      Lyn

      • Hi Lyn,
        That’s an interesting solution to the problem! I hope it works out for you.

        After having a large reel of unpigmented PLA sitting out on a spool in my studio for over a year now, I’m not noticing any problems with absorbed water while printing. And that’s living through our subtropical summer humidity. So I’m not bothering with any sort of moisture control at present.

        Do note that 58 degrees Celsius is pretty close to the 60 degree Celsius that is often quoted as the beginning of the softening zone for PLA. So you wouldn’t want to go much over that temperature for any length of time, or the filament might start to sag or stick to itself.

        • Lyn says:

          I had a reel of clear filament that worked the day that I got it but a few days later the prints were super light and soft. After 4 hours in the dehydrator the prints were firm but still exhibited a few small areas that were soft. I made no effort to loosen the filament on the reel so I’m attributing that to the problem. I’m still experimenting. Super humid today after almost 11 inches yesterday.

          • I guess maybe it is down to the quality of the filament. Perhaps some blends are more hygroscopic than others. My clear roll came from MakerGear years ago. I’ve heard some horror stories about cheap filament from shoddy Chinese manufacturers. Maybe you got a bad batch.

  4. Loong Jin says:

    I just tried the oven-baking tip at 100°C (the lowest this oven would go) for 10 minutes with a small amount of red PLA from InkBox3D. It seems to have worked — perviously, it would jam the moment my feedrate dropped below 50mm/s (0.2mm layer height), and intermittently at speeds above that. After baking, I’ve had two flawless prints, one with a layer height of 0.2mm, and another with a layer height at 0.1mm, with some portions going as slow as 20mm/s, and no jamming! \o/

Please leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s