Investigating different methods of filament drying (Dehydrator, Vacuum, Oven & Desiccant)

I recently bought myself a vacuum chamber that I want to use for degassing resin and vacuum casting. I thought this might also be a great new method to dry my filaments without harming them by high temperatures. Well, turn’s out, it, unfortunately, isn’t that simple and I’ll show you why in a bit. Usually, we distinguish between surface and core moisture. You can get surface moisture if you, for example, get a roll of material out of your air-conditioned house into your garage on a humid day. Moisture condenses on the material and cause problems even with non-hygroscopic material. This moisture, if it’s not soaked up by the material will quickly evaporate in normal ambient air and cause no longterm problems. Core moisture on the other hand is moisture in the material itself, and that’s what we’re usually concerned about because to get rid of it, it needs to diffuse to the outside and then evaporate there. We assist evaporation usually by heating air, which reduces the relative humidity of it. We can only dry until we reach a moisture equilibrium between the drying medium and the moist medium, which is the reason why warmed up and, therefore, dryer air can remove more moisture than cooler air. The dew point of the air characterizes this. The lower the drier and the better it will be suited for drying. Higher temperatures probably also increase the diffusion speed to transport the moisture from the inside to the outside of the material.

Surface moisture VS core moisture

Surface moisture VS core moisture

A vacuum reduces the boiling point of the water and might, therefore, help when trying to reduce the moisture content because it’s easier for the moisture to get out of the material. In industrial injection molding, it’s totally common that the plastic pellets are dried before they enter the process. Dry material is essential to ensure process stability, appearance, and mechanical properties of the product. In our 3D printers, we also melt the plastic, and moisture in the material is a factor that can often decide whether a print is successful or not. Many overlook that factor because at first glance, common materials like PLA, PETG and ABS don’t seem to be affected that much. More technical materials like Nylon or others behave way differently when not properly conditioned and will bubble up, string and just break apart after printing. But also our common materials are affected by water that is contained within the material. Moist or just not ideally dry material strings more, oozes more and in the case of PETG for example can lead to severe hydrolysis where the water in the material causes the long polymer chains to break up which causes the material to become brittle.

Samples printed from wet PA12 (Nylon)

Samples printed from wet PA12 (Nylon)

I also store many of my materials simply on a filament rack, which works okay most of the time. I only store my really hygroscopic materials a big plastics box with a ton of desiccant on the inside to keep it dry. Since even short exposures to normal outside humidity can cause moisture pickup and degrade the material during the process, I even have one box out of which I can directly print. If you also want to build one on your own, then check out the video I made it.

A couple of months ago I bought myself one of these generic single stage vacuum pumps on Amazon with a large pot-like container and an acrylic lid.  Not the cheapest but definitely, also not the most professional one but probably well suitable for degassing resin and some vacuum casting that I would like to do in the future. One of the things I just had to  do at first, was to get a cup of cold water to a boil and it’s able to do that in around a minute. Great! So, the idea was that if I can get water to boil at room temperature, can I also get the moisture in filament to boil out, without the need to heat it. A food dehydrator or your kitchen oven work for that process as well, but not only do they both consume considerable amounts of energy, the heat can also damage the filament itself. Obviously, if you overheat the material, for example, above the glass transition temperature, it can become soft and deform on the roll, or as some others have already experienced in the past, you can even melt the roll it’s spooled on to. Even though you might have selected the right temperature, you can still damage the material and over-dry it. This means that the additives, that for example make the material less brittle, can be damaged at longer heat exposure times, or the polymer can even oxidize. When that happens is hard to say and depends on material and manufacturer, but the less exposure to heat the materials have, the better.

I loaded the container with a roll of PLA, TPU, ASA and ABS filament, ran the vacuum pump for 15 minutes and then shut all the valves and let it sit. I wasn’t able to run the pump for longer because it would otherwise overheat. Though after 15 minutes, the vacuum limit for my setup was reached, and the dial didn’t move anymore. I regularly removed the vacuum, weight the material on a sufficiently precise scale that I calibrated before each use to mitigate temperature drift. At first, I measured in shorter intervals, later they got longer and altogether I performed regular measurements for 10 days. Looking at the results, we can nicely see that at first we remove a higher amount of moisture from the materials as after a while. At around the 5 day mark sitting in the chamber with regular evacuation, the weight stagnates, and we don’t remove any significant amount of moisture anymore. I plotted the results in two different ways. Once, just the weight itself, but depending on the initial weight of the material 1g of moisture can be siginifant or also not. The second plot shows the moisture change rate in percent and nicely shows when nothing is happening anymore. Even though we removed one to two grams of moisture from every roll, this is still only a reduction by around 0.1 to 0.2%, which isn’t a lot. Out of curiosity, I also printed a pair of 3DBenchys from the PETG. Still, I couldn’t really see a significant difference in print quality, which poses the question if the vacuum was able to remove all of the moisture or only some.

vacDry_all.png
vacDray_changeRate.png

So I left the materials outside to re-hydrate and also weighted the material regularly. Interestingly the moisture uptake is again high at first and then slows down until it gets to a more or less constant level after five days. Almost all the moisture we removed during vacuum drying seems to be back in the material which shows that if you want to keep your material dry, you also need to store it properly.

hydrating_all.png
hydrating_rate.png

Next I wanted to compare vacuum drying to heat drying in a food dehydrator that I’ve been using in the past to condition my materials.  This test should give us an indication of how long this method takes to dry filament and if it can remove more moisture in comparison to vacuum. I stacked the material in a way that ASA was on the bottom and PLA was on the top so that I could raise the temperatures a little more because the heat comes out of the bottom. With a setting of 75°C I get 65°C for the ASA and around 55°C for the PLA, which reduces the risk to soften the materials but to be high enough to dry them properly.  As a rule of thumb, stay at least a couple of degrees below the glass transition temperatures of the materials you treat. You can usually find that value in the technical datasheets. I newly added a carbon fiber nylon because that should be especially critical for moisture and this roll has been sitting in my office unprotected for a year now. We again see that at first, the amount of moisture that is removed is high and slowly decreases and stabilizes. After 12h, there is barely any change any more. In the past, I usually let it run overnight, so this test now confirms my usual procedure. Still, if you really want to know how long you need to dry your specific materials with your setup, I’d strongly recommend getting a decently precise scale. During drying, weight the rolls from time to time, until there is barely any change anymore measurable. Comparing the two methods shows that heat drying is not only the quicker way of drying, but also it removes more moisture in total. So I’d say that at least with one of these generic single stage vacuum pumps, vacuum drying is not as efficient as drying using warm and dry air.

dehydrator_all.png
dehydrator_rate.png
compare1.png

Maybe combining vacuum and medium heat might yield better results though I wasn’t able to try that yet. What I did try was vacuum drying and storing a couple of materials in the chamber with additional desiccant. The silica gel took up quite some moisture, which also reduces the increase in pressure caused by the evaporating water. This circumstance probably helped to get a bit more moisture out of the materials, though it still wasn’t as effective as heat drying. At the current point, I’d suggest using this method for storing demanding materials but not for the initial drying.

vacPlusDesicc.png

Next, I wanted to work on the question if the filament spools are dried more on the outside than on the inside due to the lack of airflow, especially with these perfect winding spools. For this reason, I stored the dried spools of material in one of my dry-boxes and under vacuum for a week. The idea was that after a while, the moisture might equalize again within the spool. I then dried it in my food dehydrator for a second time. The interesting thing that I was able to see was that some materials seem to pick up, and some seem to lose a small amount of weight while being stored. This might be moisture from the atmosphere within the drying container, or just pick up during the time, I had them in ambient air while weighing. Even though noticeable, the change is around one order of magnitude lower than what we lost during initial drying. The drying step after storing again removes only a tiny amount of weight from the rolls. For ASA, PLA and PETG only in the range of what they gained during storing. Only the CF Nylon lost a more significant amount again. Since this spool was very loosely wound, I don’t think that it was due to unequal drying in the first run, but this probably shows that more demanding or more hygroscopic materials might require more conditioning than others. Since the other spools only lost tiny amounts of moisture, I think we can assume that drying in a food dehydrator for around half a day equally removes moisture from the whole spool and not only the outer layers. Even just 4 to 6 hours usually does the trick for PLA and PETG.

equalDrying.png

Just on a side note: Out of curiosity, I also threw two rolls of undried PETG and PLA in the silica filled drybox and those lost quite a significant amount of moisture during the week. Additional drying in my food dehydrator only removed a little more. So a properly sealed box with plenty of silica in it, which I for example buy by the kilo in buckets might already be a good start to keep your less demanding filaments in a good state. Due to health concerns, I don’t recommend desiccant with indicator anymore and rather suggest you put a hydrometer in there. As soon as the level goes above 10% regenerate the desiccant in your oven at 110°C for 2h.

storeInDesiccant_summary.png

Since we’ve seen that the Carbon Fiber reinforced Nylon didn’t seem to dry in the food dehydrator properly, I tossed that roll, together with three more rolls of plain and carbon fiber reinforced Nylon into my kitchen circulation oven. I used real 85°C instead of the 65°C of the dehydrator to find out how much more moisture the higher temperatures can remove and in what time-frame. I opened the door at regular intervals to let the moist warm air out to improve drying. The results show that even though the spools were pre-dried, the elevated temperatures were able to remove significant additional amounts of moisture, and even after almost a whole day, I was still able to measure a small decrease. That lets me conclude that if you work with higher technical materials, it is necessary to find a proper drying solution and check the weight to see if the process finished. Even though I’ve successfully been using food dehydrators for many of my materials, at least mine is not sufficient for some of my more demanding filaments. If you have questions about drying temperatures and times, also don’t hesitate to contact the material manufacturers.

ovendrying_all.png
ovendrying_rate.png

So let‘s quickly summarize the results:

  1. The importance of dry 3D printing filament is too often overlooked! If you have problems printing a material, make sure it’s sufficiently dry!

  2. Vacuum drying works, though it’s not as efficient and effective as drying with heat.

  3. Adding dessicant to the vacuum chamber helps a little.

  4. A food dehydrator is sufficient for common materials like PLA, PETG and ABS though struggles with higher demanding polymers.

  5. Weigh your material regularly to determine when drying is finished if you’re in doubt.

  6. Storing material with sufficient amounts of dessicant does not only keep common materials dry but even removes some moisture.

  7. Storing materials in a vacuum chamber with dessicant is even suitable for highly demanding materials.

Stefan Hermann