I’m sorry but that can’t possibly be true, there is no way this process results in knowing the percent moisture in the filament.
If you take the desiccant out of this equation the filament will reach an equilibrium with the air at a given temperature, but the air RH from this process isn’t the percentage of moisture in the filament either. It’s not even the percentage of moisture in the air.
There are scientific instruments that accurately measure the moisture content of plastics. However they tend to cost many thousands of dollars. Wood can absorb large amounts of water, which makes it easy to detect and moisture meters tend to be inexpensive. However plastics absorb very little water — only minuscule amounts. Thus, the instruments need to be extremely accurate at the low end of the scale. When I look at the material science literature I see that PLA and PETG are characterized as being hydrophobic. That they don’t absorb water but rather that there are airgaps in the material where moisture can seep into.
When I have seen the graphs using these highly sensitive instruments, PLA soaked in water will immediately begin giving up its water when moved to a dryer environment and will tend towards equilibrium with the surrounding air. That is why those cheap hygrometers work, because the moisture in PLA tends to be in equilibrium with the surrounding air in a sealed container. Likewise with desiccant as it removes moisture from the air which PLA will then tend towards equilibrium. The air filtration in my house tends to keep the relative humidity at around 36%. I have never had a problem with wet filament even though I have left spools sitting on the AMS lite of my A1 for months at a time.
Polyamide is a whole different beast.
Yes you can (depending on how you define “overdry”). It’s basically what I do. But the way I dry filament is very different from how most do it but it’s close to your example of 80C/0%.
Most don’t or simply cannot “overdry” their filament due to limitations with drying.
It’s cumbersome to explain but I started with a standard Sunlu S2+ dryer. I put a spool in after weighing and monitored weight and humidity while the filament “dried”. Eventually the weight and humidity stopped changing which most take as “dry”.
I then turned on the dry air purge and humidity and weight started dropping again. And there is the rub. After “drying” to constant weight and humidity, it turned out the filament wasn’t actually completely dry.
Any drying method that uses ambient air will be somewhat limited by the ambient humidity in that air but using very dry air is indeed how you can “overdry” filament.
But if it matters or not really depends. I have a problem with high ambient humdity and had problem start appearing in my prints as my filament picked up water. I had to do something to get print quality back.
Filament only needs to be dry enough to print well and many can do that with food dehydrators and standard dryers. But those will leave some amount of water still in the filament. To be completely accurate, my dry air method also doesn’t/can’t remove all the moisture in the filament either but it removes more than other methods.
thank you for all the information! you’ve done a LOT of research! appreciate all the responses from everyone.
So all those museums that do it this way are wrong? I suggest you go read up on hydrodynamics on how waster moves, then it may become clearer what I am saying. I also did not state absolute moisture content, but a close understanding of moisture content.
It’s a minor misconception. Museums and archival storage folks do use humidity targets for how they control environments - absolutely true. The reason is humidity is easy. For example, say they’ve figured out parchment lasts best at 20% humidity (I don’t know what humidity is best - just for example). That’s then how they make it last - keep the humidity at 20%.
That is very different from moisture content even though and humidity and moisture content are related through the physical properties and composition of a material. That relationship is fixed. At 20% humidity, for example, one material may have a water content of 30g/kg or whatever.
Here are curves that relate the water content of various desiccants to humidity. Not only are they all different, none are straight lines.
But, since the relationship between water content and humidity is fixed for a material, you can use relative humidity instead of moisture content - like the museum and archival industries do. But it isn’t actual moisture content. For us it’s easy to measure humidity but not so easy to measure moisture content. And we don’t know the shapes of the curves for our filaments. But since the relationship between water and humidity is fixed for a material, we can just use humidity.
Also, you’re right about how the water moves. 4 spools of filament in an AMS with desiccant is a complex environment and the humidity reading is more a comment on desiccant than filament thanks to filament being a poor desiccant. Multiple water sinks and sources gets messy.
Minor amplification but important, the relationship is fixed at a particular temperature.
Very true! Anyone comparing humidity values should make sure they are taken at the same temperature or corrected to the same temperature.
In practice, the way I do it, I’m monitoring drying progress using the hygrometer in my filament dryer. PLA is at 55C, PETG is at 65C (going by memory so might be off). Anyway, I’m always making the “while drying” measurements at the same temperature for that kind of filament.
My in-storage humidities are all at room temperature around 25-26C. It’s easy to get sloppy and leave that off but it’s there behind the scenes and absolutely matters. 
But that has nothing to do with water content of the filament. You’d have to examine it like a fiber optic end face.
The humidity inside a sealed container with a spool of filament has everything to do with the water content.
Why do you say it doesn’t? Please explain but know it will need to be a good one because an entire industry disagrees.
Do you think graphs like this are bogus? Or do you disagree that filament itself behaves like a desiccant?
I think we can agree that we are basically on the same track, just saying it differently. A really wet spool in a closed container with a meter will show that eventually the air will become saturated to the level of moisture in the spool, at a given temperature. The warmer the air, the more moisture it can hold.
I also never claimed this will tell you exactly how much moisture a spool holds, but a close Idea of content. The spools are packed too tightly to fully release all moisture. Over time, the moisture will migrate to the rest of the spool from the middle. Heating it, and having a means to remove the moisture from the surrounding air is the only true way to draw out the inner moisture as it will migrate from the cooler middle to the warmer exterior of the filament roll, where it evaporates, leaving the filament. This is what a filament Dryer does, heats the air surrounding the spool so it can absorb more moisture, circulating it with fresh cooler air, while slowly warming the filament from the outside to draw inner moisture out.
Desiccant in a sealed box does the exact same thing, but without heat, just takes longer, it absorbs the moisture in the air inside the sealed container, the dry air now absorbs moisture from the spool, which in turn gets absorbed by the desiccant until it reaches its capacity, at which point the air in the container will absorb an equal amount of moisture from the filament that it contains. It usually takes several days depending on the temperature. Keeping in mind that moisture seeks equilibrium, when the outer strands exposed to the air reach a lower level than the inner strands, those inner strands will release moisture to seek the same level as the outer strands, until equilibrium is reached.
One other factor involved is that moisture flows from cold to hot, a cold spool inserted into a warm container will release it’s moisture quickly into the warm air, on the flip side, if you pull a warm spool out of a dryer into cold damp air, it will act as a sponge and start soaking up that moisture for as long as it sits out in the open. ← Reason I bag as soon as it is removed from an AMS, 66* basement with RH of 60% or more.
Study ERV’s that are used in HVAC, then you will get a better understanding of the way this works. These are systems designed to transfer fresh air with stale air while blocking the moisture from mixing.
I think you’re conflating two separate things: 1. Filament moisture, 2. Container moisture.
I am saying there is no non-destructive consumer equipment to measure the water content of filament. The water content of filament can’t be assumed by measuring air moisture content inside a container. This is why the accepted answer is correct.
There are some misconceptions here though. If you look at micrographs of filament the surface is actually very rough with pits and cracks. Even if strands are tightly and closely wound, there are still huge open areas between strands for water to get out. A water molecule is extremely tiny compared to surface roughness. There is a barrier but it’s not that big.
Something else is collisions and speed of the gas molecules. At room temperature and atmospheric pressure, water molecules are moving fast. Going by memory but if you ask a search engine, it’s something like 800mph but hitting other molecules every micrometer or so. There’s lots of movement we can’t see but it’s there and why it’s no feat for water to get out.
I don’t know this but I think the bigger water concentration gradient is in the filament itself from outer surface to inner core.
Another factor is heat/temperature. There are energies associated with each water molecule stuck down on filament. Some sites hold water stronger than others. Without additional heat above room temperature, some water will not let go so some water can be removed at room temperature but some cannot.
You can find out more details if you look up heat of adsorption.
Not arguing against this, what I meant is that the inner layers of the tightly wound spool will take longer to fully release the moisture than the outer exposed layers, which in turn means it takes longer to fully dry a spool that has a high % of moisture than one that is loosely wound as it takes longer for it to warm up to release the moisture as opposed to migrating the moisture.
This is why some prints will show great layer lines and towards the end of the print when half the spool has been consumed you start seeing bubbles and moisture related issues. IOW the spool wasn’t dried long enough and let stabilize before printing with it.
While I am talking about measuring the humidity of the air in a container and using that as a proxy for filament moisture, it’s completely valid.
If it wasn’t valid, desiccant wouldn’t work.
And there is lots of consumer grade equipment to measure water in things like drywall and wood, and there may be none specifically for filament, the consumer stuff uses RF and conductivity. It might work with filament but I don’t know how well or accurate.
But there absolutely is a way to measure filament moisture using a simple hygrometer. Have you even tried it?
Easy to test. Open a ziplock bag to atmosphere and drop in a hygrometer. Let the reading stabilize and write it down. Then add a spool of filament and seal it back up.
If you do open storage your filament will have already equilibrated with room air so don’t use one of those. Either take a spool from your dryer or AMS that hopefully is a bit more dry and put it in the bag with the desiccant. The reading should change.
The reason it should change is the filament either gives up or absorbs water depending on if its equilibrium humidity is above or below the humidity of the air.
And you should also know others are using this technique now and not one has said it doesn’t work. It really does. Please try it for yourself or bone up on the desiccant industry literature on how this all works — because it really does. I use it just about every day. No inconsistency. No bad answers. Just a reliable way to get an indirect measure of filament moisture. Emphasis on indirect.
Dude, don’t just downvote. If you disagree with me, explain where what I say is wrong. Don’t just interject “it doesn’t work” or other pat comments.
If it doesn’t work, what is the physical basis for it not to work?
I’d really like to know if I am indeed wrong. I haven’t just taken a position for the hades of it. I have a background in this stuff and have done lots of additional reading of desiccant industry papers and articles.
First you should know humidity in a closed container is exactly how the desiccant industry tests their desiccants. It’s also how they prepare their desiccants. You can buy silica gel pre-packaged in 5% RH increments out of on the shelf stock. They sell it to the museum and archival storage industries so it can be used to help preserve different items. If they want to preserve some document at 20% RH because they last the best at that humidity, they buy 20% RH beads and put them in the display case. But you can also buy beads in 1%RH increments special order.
They make those measurements all the time using relative humidity. So if it works so well with desiccant testing and prep, why is it invalid when applied to filament? The fact is filament also has desiccant properties and can be treated like a desiccant itself. Everything they do in the desiccant industry can be applied to filament. The biggest difference is we just want dry filament. At least as of yet, there hasn’t seemed to be a need for filament at some nonzero humidity. We just go for as dry as we can get it, but it’s the same processes and behaviors as a billion dollar industry.
So please don’t just downvote. Anybody can do that. Explain why I’m wrong.
I agree with you there but this really has nothing to do with my overall point. You’ve added other complexities to the issue which are valid but muddy the water.
Where I see this is on spools that have dried and leaked up to some extent while in storage. The behavior during redrying is slightly different from a “new” spool in that the dry is significantly faster. My interpretation of that is the new water stuck down nearer the surface of the filament strands and so can get back out easier. This fits with what you say and I think is direct evidence of it being true.
But in all my discussions, I have intentionally left off those kinds of things because they are nuances to the major theme - using RH to judge moisture content.
It’s been my opinion for a while that when you get near dividing lines between moisture levels that still print ok and those that don’t, you start seeing the effects of doors and fans and one side of a print can look better than another. When splitting hairs and near the edge like that I can certainly see a print starting good and ending bad.
But the solution there is not to rush drying. Short dries do risk not getting water liberated from the center of a spool because it hasn’t gotten to temperature. And there is still some restriction to water getting out from the core once freed from being stuck down. Those both have to do with technique instead of the principles it’s based on being bad.
If people don’t rush a dry where the entire spool has a chance to come up to drying temperature and then wait long enough for the water to get out, they aren’t issues and don’t invalidate the larger point.
I use the hygrometer in my dryer to decide when my drys are finished. Humidity initially spikes during my dries but comes down after. All I do is wait until the humidity has hit my target and I’m done. I pull at 19% RH at drying temperature where the water still coming off the spool has really slowed down.
I have yet to have any issues whatsoever from the inner part of a spool showing water effects. If anyone is seeing that, what it’s telling them is they aren’t giving enough time for the entirety of the spool to get to temperature and not giving enough time for the water to migrate out. It’s operator error. We can’t easily change the issues with a torturous path out. We just have to give enough time.
You aren’t so outspoken now, are you?
You keep just downvoting. I swear I am able to hear other sides but you need to back up why you think I’m wrong. It’s not helping your case to ignore the facts and just hit my posts with downvotes.
If you can justify your position, please do. But at this point I don’t think you can. So until you are willing to actually discuss this stuff intelligently, downvote away my friend. But it just makes you look silly and unable to justify your opinion.
Measuring actual water content in a polymer means Karl Fischer titration or loss-on-drying (halogen/TGA). Both are lab methods and both are destructive, so neither works as an onboard sensor feeding an AMS mid-print.
The non-destructive alternative is equilibrium RH, which is real and does track how wet the filament is. But it only reads at equilibrium, so you’d have to seal the spool and wait hours for the air to settle before you got a number, which is the exact wait you’re trying to skip. And it gives you moisture state, not a water-content value you could convert into retraction/wipe numbers without a per-material calibration curve.
So the concept isn’t crazy, the physics just don’t cooperate. Real-time water content on a moving spool is a lab problem, not a $20 sensor.
