Measuring filament moisture

Other topics
1

Note - @krellboy has a good point and there is some misconception in this post. See down thread for details but take this post with some salt…

I’ve been saying over and over that humidity can be used as a stand-in for moisture content and can be used to decide if filament is dry enough for printing or not. Not everyone believes me.

Turns out the plastics industry has the same problems we do and they already solved how to deal with it. Spoiler - they measure the humidity of air around the pellets to determine moisture content. It works.

This unit doesn’t measure humidity directly. What it does is convert the water in the air to hydrogen and measures that so they can have answers quick. Hygrometers respond relatively slowly.

But they are indirectly measuring the humidity. That is where the hydrogen comes from that they do measure. There’s just a chemical conversion step in there. But what they are ultimately measuring is the humidity surrounding the plastic to determine water content of the plastic. And they claim it’s as accurate as a Karl-Fischer titration (the gold standard in moisture).

They aren’t measuring the water vapor directly but the hydrogen evolved from water vapor from a chemical reaction. Bam! It’s the same thing with a minor technical difference that gives them results faster.

The HydroTracer 3 sells between $14,000 and $23,000 depending on accessories. If you don’t mind waiting 30-45 minutes you can do the same testing on filaments for about $10 for a Praki thick-walled sealing container and a cheap hygrometer. It might not be as accurate and certainly is slower, but the principle is the same. Humidity tells you moisture content. They just go a step farther to convert to actual moisture content and get the numbers faster.

From their web page - it’s a schematic of how their instrument operates. They let the plastic equilibrate with the air and then indirectly measure humidity. But make no mistake - the underlying principle is using humidity to know moisture content.

IMG_4781
IMG_47811357×1262 170 KB

So for those who doubt it works, companies drop thousands of dollars on testing equipment that does the same measurement albeit much faster. But we can do it on the cheap if you don’t mind the 30-45 minutes you have to wait for a hygrometer to settle. You can even use a ziplock bag instead of a Praki. The plastic bags are just subject to water vapor diffusing across the bag walls and affecting the measurement if you leave them longer than necessary.

And they also appreciate how controlling water content (not just minimizing it) is key to good products, strength, warping, and surface issues.

They convert back to actual water content because they can. We can’t because manufacturers don’t provide the information we need. But it doesn’t matter. Humidity value proportional to water content or actual water content. Same thing.

2

If I understand it correctly, it seem to work by heating the water out of the sample and measuring the quantity of water liberated via a stoichiometric intermediate reaction.

If someone asks “how is this different than loss-on-drying” this method is highly selective for moisture and ignores VOCs.

I’m not sure this device is the comparison you were looking for.

3

It’s exactly the comparison I wasn’t actually looking for. Just found it looking up something else.

They are doing a controlled heating to help liberate water. That’s what filament dryers do. The heat just helps evolve the water. Gets them bigger numbers and better signal to noise. Their heating is going to be carefully controlled and reproducible and really will only serve to put more water into the vapor phase.

Yes, they will be working at higher humidities because presumably water will come off faster than the heat reduces relative humidity, but they aren’t measuring the humidity or the relative humidity. They are quantitatively measuring hydrogen liberated from water.

And the part about VOCs is moot. The humidity measurement doesn’t see those either and it’s water that causes the problems. Nobody has seen the need to create a filament VOC remover. They are background.

Technically it has differences but practically it is the same thing.

4

Maybe we are talking about different things. I thought you were using this device to validate people using equilibrium humidity as a moisture estimation technique.

5

I am.

They heat the sample. Equilibrium humidity is at room temperature because of the temperature dependence on relative humidity. The difference is they evolve more water faster but it’s still just water. You could do the equilibrium humidity test at elevated temperature too but you’d need to make sure all tests were done the same to compare results. Even saying room temperature allows for differences and errors to creep in but they aren’t even measuring minor. We don’t need super-accurate numbers for our purposes. No temperature controlled oven is needed and the measurement is available to lots more people since you don’t need much hardware.

They use a chemical reaction to convert water to hydrogen and measure that. Equilibrium humidity is measuring the water itself.

These differences don’t change using water content/humidity as an indirect measure of water content.

Another difference is the measurement. I’m sure they integrate all the hydrogen evolved over the duration of the experiment. It’s more like a measuring cup if you want to think of it like that.

The relative humidity measurement is a point in time measurement. Nothing is being integrated over the duration of the test. It’s just a value. It could be integrated over the test but there’s no need.

So yeah, they heat above ambient and probably integrate but that doesn’t change the fact they are measuring water vapor. Room temperature is still heated too. Anything above absolute zero is basically heated. It’s just a matter of degree.

Also checked the temperatures they heat to and the range is 50-210C. It’s not going to be a decomposition. It may soften or even melt, but not enough information to know the strategy and I haven’t found what temperatures are used with which plastics. But 210C is extruder range. It seems the heating is likely more to speed up the test and results.

6

I’ve been thinking about it Krell and you have a point. There is a fundamental difference in the techniques. Worst case comparison is if they heat it hot enough and long enough to quantitatively recover all (almost all) of the water trapped in the filament.

A humidity measurement just relies on some amount of water proportional to the water contained in the filament to be present in the air around it. Back to the measuring cup holding all the water vs looking at something physically related to the amount of water present.

So yes, in that respect the measurements are very different and it’s obvious that if you can capture all the water even if you change it to hydrogen and measure that, you are capturing and quantifying all the water.

However, filament is a desiccant too and the humidity around it is physically and directly related to moisture content just like it is with other desiccants that reversibly adsorb water. To measure the RH in an equilibrium test is also a measurement - indirect - of the filament moisture.

So in that respect, I’ll give that the two tests are less comparable than I originally said, but they are still equivalent in many ways. Just as the instrument isn’t actually measuring water but is instead measuring hydrogen derived from the water, using RH isn’t actually measuring water content directly either. But both give numbers physically related to the amount of water present in the filament.

So good point but both tests seem absolutely valid and still similar even though the operating principle is different. How different depends on how much they heat the plastic and if they get all the water or if there is some factor involved for leaving some amount of water behind. That would move the instrument test closer to raw values that are physically proportional to the amount of water but that are not the true total amount of water.