Please can I have some advice with what I am doing wrong with my PETG HF. I have dried the it for over 20 hours between 60 and 70 degrees but I am not seeing any improvement in the print quality, I am using a P1S with a super tack plate. I have also tried printing with a grey reel which I have also tried drying but I am still getting really bad Benchy prints. The edges seem to be curling which is causing knock on effects and any bridging seems to droop a lot and the chimney is just a mess. Do I need to dry it more or am I doing something else wrong. This is my first time printing with PETG and PETG HF. I have been printing with PLA and ABS with no issues. Any help is greatly appreciated.
Please provide photos! They really help.
20h.drying at those temps is plenty. But PETG is very different from PLA. It loves to stick to the nozzle (so avoud crossing surfaces and infill patterns, use gyrpid, honeycomb, monotonic,…) and it is prone to warping/curling. Also, the classic Bambu PETG did not print well with its default settings. Generic PETG settings were much better.
Also, in my personal experience, PETG reacts positively to slowing down. I use <100mm/s for PETG, <120mm/s for PETG HF (but the latter may be conservative) with accelerations set 25% of defaults (important for small details).
So I can recommend daring to go slow. It is much easier to speed up once the desired quality has been achieved rather than achieving quality at a given speed.
Oh, and you may want to try the Smooth Plate for PETG with glue stick. The higher temp helps against warping and curling.
1 Like
Thank you, I will give that a go.
Those pics are also helpful 
Classic PETG overhang issues on the bow => Try using Generic PETG cooling settings (!) and dare to go slower.
Chimney melting: May be helped sufficiently by the above. If not: Either print 2 benchies simultaneously or enforce a prime tower by enabling a timelapse with a seemi gly static print head.
There are more tricks and tweaks with PETG (such as reduced layer height against warping and curling or keeping the chamber warm while at the same time avoiding clogging) but that is for special curling probs, multi-material printing and ironing. So not one of the key issues seen on the benchies.
1 Like
I have just tried printing a Benchy with these settings and it is much better. Thank you!
1 Like
Due to poor planning on my part and a rush job needed today, I got a great example of the subtleties of PETG HF last night.
I took a photo this morning still on the build plate. The part was printed from two spools because I didn’t have a full spool to print it in one go and it was a race to get the 2nd spool dry in time for when the running spool ran out.
Because of that, the bottom 2/3 of the print was done using PETG HF I dried in my setup to a 19% RH in the dryer. The top 1/3 was printed from the not quite dry refill spool. I only got the refill spool to 21% RH in the filament dryer by the time I had to pull it. The spool was hot when it went into the AMS and I had turned up my air flow into the dryer to full so it wasn’t as slow a cook as I normally do.
You don’t see this just looking at the part. The white hides it. But with the right lighting it really shows the difference. Where the spool was changed is marked by the little filament hair up and to the right of the hole center left. The change is pretty obvious though, and you don’t need the hair to spot it. The upper part wasn’t dried as well as the lower part but it wasn’t that different except for also trying to dry fast. However you want to interpret it, definite difference in print quality from not getting it quite as dry as it should have been.
4 Likes
Did a quick and dirty air dryer using an inline water filter canister ($10 amazon) and tiny USB air pump ($9 Amason), with another $5 of various plumbing. I will clean it up (it is far from perfect), and eventually test it with an air compressor as the air source (with a regulator turned WAY down and probably a overly good (lab grade) needle valve - it is in my spare parts bin at the moment.
Anyway, my results… Normally my house humidity is rediculus (60-70% is not uncommon), but yesterday was one of the lowest days of the year - 40%. I plumbed my dry air line into one of the filament holes in a Polymaker dryer. In 30 minutes, it reduced the humidity percentage by about 1/3 (it is a very low power air pump). I dried for 4hrs, and by the end I got down to I reduced the humidity percentage another 1/3 (from 40%, to 27% to 14%… I don’t know how this effected my filament, I did not test that. It did not solve my printing problem - but I solved that another way. It was mostly a test of the air dryer.
Without weight testing identically saturated spools, I have no way to do a real performance test - but I think it could be a beneficial way to help with drying. I can not see how the very gradual drop in humidity could hurt.
You care about the heat deflection temperature, not the glass transition temp.
I dry my filament BL PETG-HF in an oven at 74C with zero issues. Not sure how accurate the temp probe is but the temp feels about right.
Maybe YOU care about the HDT rather than the Tg, but anyone who doesn’t want their prints to deform and lose shape while in a hot environment cares about the Tg. HDT deals with how the material performs when under a load whereas Tg deals with how the material performs when in a hot environment.
A ChatGPT summary rounds it out like this:
In the context of 3D printing “heat deflection temperature” (HDT) refers to the temperature at which a material starts to deform under a load, while “glass transition temperature” (Tg) indicates the temperature at which a material changes from a rigid glassy state to a softer, rubbery state, essentially signifying the point where it begins to lose its shape under heat; although related, HDT is a measure of mechanical resistance under load, while Tg is a purely physical property describing the material’s state change.
- Function:
HDT tells you how much heat a material can withstand before it starts to bend or deform under pressure, while Tg tells you the temperature at which a material transitions from a rigid to a soft state, regardless of load.
- Application:
When choosing a 3D printing filament for a specific application, HDT is more important if the printed part will experience significant stress at elevated temperatures, while Tg is key if the material needs to maintain its shape at a certain temperature range without external load.
Clearly in the context of heat exposure only - such as a print that will be left inside a parked car, or a self-printed filament spool inside a hot dryer - Tg is what matters (note the part where the summary explains that Tg is “essentially signifying the point where it begins to lose its shape under heat”).
Late to the conversation, but for anyone reading this in the future… the idea of pre-drying the air would seem to be of little actual benefit. Simply heating air greatly reduces its relative humidity, since hot air can hold more moisture than cold air.
Starting with room air at 68F and 50% humidity, simply warming it up to PETG drying temperature of 65C (149F) drops its relative humidity to just under 5%. Pre-drying it would make that lower still, but I think it is way into the realm of diminishing returns.
1 Like
While that is true that heating air reduces RH, my testing has shown that residual moisture still matters. If you read through the Filament Drying Preliminary Results thread, I tried using just room air to purge my S2 filament dryer. Performance/time to “dry” and ultimate level of dryness were compromised by ambient humidity and slowed filament drying. The results were very similar to using a door prop on the S2.
When I put the air drying column back in line, weight started dropping again and RH% in the dryer started dropping again. I had let weight and humidity plateau before adding the drying column. Again, as soon as I put the desiccant back in line, weights started dropping again and so did humidity in the dryer. I saw it with my own eyes.
I really don’t mind you guys dropping in and posting your “feels” about filament drying, but if you are going to make these claims it is incumbent on you to back them up with actual data. That is how science works. You can doubt my results all day long but until you can show where I went wrong and you got better results it’s just hot air. I’m not saying you’re wrong. I’m just saying my data with my setup does not support your conclusion.
I’m not in any way saying your hypothesis is incorrect, but I have actual experimental results and over and over, and as time passes I get more and more evidence that fits and supports my claims. Dry air does matter and it has greatly improved my print quality. The S2 alone, with ambient air purge and with the door propped open, stopped early and left more water in the filament.
Please do the experiments and report your results. The Filament Drying Preliminary Results thread would be a great place or start your own. Without results and data, however, and an explanation why dry air doesn’t make a difference, data and documented results are a bit more believable.
But maybe I made a mistake. If so, find it. Let’s see actual evidence and scientific explanation why I didn’t see what I saw.
1 Like
I might see where everyone keeps getting led wrong by relative humidity.
Take a trash bag of air at some relative humidity and heat it. The relative humidity changes (lowers). The actual amount of water in the bag does not change. If you cool it back to what it was before, the relative humidity will return to the previous value.
All relative humidity is is a measure of how much water a gas (air) is holding vs how much it can hold at that temperature before it starts condensing out. Heating the air does not affect how much water is in the air or the ratios of one gas to another. It’s all still there just as it was before.
Diffusion is a lot more complicated but that’s a lot of what carries moisture away from drying filament. There are online calculators for molecular and atomic velocities and mean free path for different sizes and masses of molecules. They only depend on molecular/atomic weight, pressure, and temperature. What you’ll find is water molecules at room temperature move like rifle bullets with speeds around 2000ft/sec but hit other molecules every few nanometers or so. If you look up random walk and diffusion, there’s lots more information.
As you heat a gas the molecules and atoms move farther apart as the collisions get more energetic. The ideal gas law (PV=nRT) explains the volume increasing with temperature so that’s the move farther apart bit. Mean free paths will get longer. But the makeup of the gas doesn’t change. A water molecule hits the same proportion of other water molecules, oxygen molecules, nitrogen molecules, etc. That means it’s seeing essentially the same resistance (aside from the decreased density) to moving away from the filament as it would in a cooler filament dryer.
What changes with dry air is the makeup of that air. With fewer water molecules in the gas phase to run into, net flow of water away from the filament increases. In that respect it’s like any process that depends on a gradient. The movement always depends on the differences in concentrations, temperature, or whatever process is being considered. It’s why if you want to keep a CPU from overheating you cool the other side of a heat sink - it makes the heat flow faster. It’s why water flows downhill faster over a waterfall than along a stream.
I need to think of better ways to explain it but dry air does make a difference in how fast filament dries and how well. Relative humidity is only a measure of how much water air can hold before it condenses out.
To reject that is to say drying filament at 80% RH ambient would be the same as drying filament at 20% ambient RH and that is obviously false. Both have excess capacity to hold moisture. As long as you don’t evolve so much water that it starts raining in your filament dryer, why is drying at 80% humidity slow or just flat out ineffective when drying at 20% gives results?
And, if you allow that drying at 80% is slower than drying at 20%, why wouldn’t drying at 1-2% RH not be even faster.
And since I’ve opened that can of worms, the point has been raised about why a food dehydrator with a heat-reduced RH% lower than what I dry to in my filament dryer would or wouldn’t work and I think it probably would but haven’t tried it to have any data. I’m using RH% in the chamber to determine when I’ve dried sufficiently so why wouldn’t a food dehydrator be better?
There’s multiple issues there and drying is a dynamic process. One is the flow through a food dehydrator defeats using an RH level as an endpoint. All you have is time or indirect measurement of filament moisture at equilibrium in something like a poly cereal box which works very well albeit slow. Weights tell you how much water you lost and if drying was effective but doesn’t tell you anything about moisture content. @NeverDie posted a photo of a model printed with filament that was dried unevenly with clear banding of moisture defects so being consistent where you dry to can be important.
As long as you dry long enough with a dehydrator you should be good but varying moisture levels can affect calibrations so if that’s important, that may be a consideration in choosing a drying method.
2 Likes
I don’t have any inside information on the composition of PETG-HF but usually to print faster a filament has to flow better, and one achieves this by shortening the polymer chains. Unfortunately, those long molecules give you all the nice physical properties.
I’d also like to see a “tough and heat resistant PETG” but to be honest the filament makers will probably just use that as an excuse to sell off their old stocks of “slow” PETG at higher cost!
I have two degrees in mechanical engineering and I studied plastics during my engineering masters. You do care about HDT.
I literally just told you that I put finished parts and filament in my oven at 74c, well above the Tg. and they didn’t deform. If your assertion was true then all of my parts would have melted.
I’ll give an example of why you don’t care about Tg. Let’s say you left an elastomer like a rubber in a hot car - it wouldn’t permanently deform. That’s what we would expect right? But all of these elastomers have Tgs that are low, very low usually below the freezing point of water. Tg is often similar to HDT but it’s not actually what you care about when designing for a hot environment.
Tg isn’t a great way to measure a material’s heat resistance for a variety of reasons. One of the main reasons is that it only considers how “soft” the polymer gets but doesn’t take into account any additional stiffening effects of the additives. For instances, it’s well known that adding glass or carbon fibers increases a plastic’s heat resistance but if you look at the data sheets often times the same plastic without those fillers has nearly the same Tg.
You can see with your very own eyes what happened to my printed part in the original post. I also have an equally deformed battery holder for my dash cam that almost instantly melted in the parked car. Once both of these were reprinted with PETG Basic, they’ve been going fine since then.
So, I guess I’ll just tell those two pieces of deformed plastic that all the other engineers in the world are wrong because some random dude on the internet with “two engineering degrees” says they are not deformed. Do you think I somehow printed that spool holder that way?
Maybe 2 degrees is 1 too many. It seems that once people get enough degrees they start to claim that what they see with very own eyes isn’t real.
And what does your education in plastics tell you about flow rate? You need to get your money back from your university if it doesn’t tell you that flow rate is increased by shortening the polymer chain length. And then what does it tell you happens to the mechanical properties of plastics after you shorten them?
And where did you make any reference to putting finished parts in your dryer? I (as well as others all over this forum) have posted pictures showing deformed parts. Since you say this doesn’t happen according to your 2 degrees, explain the pictures.
Additionally you might want to check the TDS because Bambu themselves say the HDT is 62C - 4 degrees LOWER than the Tg.
I mention drying in my oven in my first post that you responded to.
Your spool is only melted in one corner, it looks like there’s a hot spot in your dryer. Where is the heater located in your dryer? Is there a strong air current in the dryer to evenly distribute the heat?
You mention drying your filament in the first post, not printed parts.
And I conceded in the original post that my dryer may be partly to blame. However, that does not explain the battery mount in my car. PETG HF deforms within minutes in my car whereas PETG Basic does not deform at all.
And why would you and your education not expect it to deform in a dryer at 65C. The HDT you insist is so important is 62C (unlike almost all brands of PETG Basic, which run somewhere around 80C). What would an engineer say should happen to a plastic that has a Tg of 66C and an HDT of 62C if it’s put in an enclosed heat chamber for hours on end at 74C - where you say you put yours?