Oozing, blobs, and strings on S5 / PETG

[cheuschober 1]

Hi, I've spend the last couple of weeks churning out calibration prints to no avail and thought I'd come here for ideas before I give up on PETG on the S5 entirely.

As stated, I'm getting oozing, blobs and strings with any prints I've tried.  Starting with the Engineering - Normal settings on a 0.4mm  nozzle I've tried towers and modifications to:

• temperature
• retract distance
• retract speed
• line width
• fan speed
• flow

Nothing's been a silver bullet but I'll admit that something is starting to smell "fishy" about temperature. So far, I've tried both MatterHackers Pro and Ultimaker-branded filament and both seem quite happy printing as low as 215C (I've not pushed it further) and I've had problems with PVA burning with this printer before. 215C seems really low for my best quality PETG prints to show up.

At this point, I'm curious what steps towards diagnosis would be sane or, does the S5 just do really poorly with PETG?

[gr5 2,172]

Please include photos of what you are talking about.  We can't tell if you are complaining about things that are too small/minor to be fixed, if you are on the edge of fixable or if your issue is 10X worse than normal printing.

[TimonR 21]

@cheuschober: A minor but often overlooked problem with PETG is moisture.

At lower temperatures the material doesn't have low enough viscosity to flow properly, leading to underextrusion. At higher print temperature, you basically start to degrade the material. This degradation is (very) moisture dependent.

The effects that you describe are related to degradation. Why?
Degradation means that you polymer chains get broken, leading to shorter polymers which have lower viscosity. This means the material will be more liquid-like at those printing temperatures, leading to more severe stringing and oozing.

 

My first suggestion therefore would be to dry your material in a PrintDry / Oven / food dessicator at 50-70C for a few hours / overnight. Then try to print your model again.

If this is the issue, consider purchasing a PrintDry/Polybox other storage unit to keep your material dry during printing.

Hope this helps!

 

[gr5 2,172]

@TimonR is this moisture issue also true with CPE and CPE+ filament?  Do they also have the viscosity issue caused by moisture?

[TimonR 21]

Hi @gr5,

In principle all condensation polymers (polyesters like PLA, PETG, CPE = branched PETG, CPE+ = PCTG, and polyamides/nylons) have this issue. These polymers are 'coupled' by connecting monomers, which releases water. The opposite (decoupling / hydrolysis) happens when you have moisture and high temperature.

See here a very nice explanation on viscosity decrease at higher temperatures (hydrolysis) of polyamides: http://extrusionist.com/presentations/prejudices about polyamide.html (from slide 20 onwards).

Now, it basically depends on the moisture uptake of these polymers how 'bad' the situation will be.
CPE is a branched PETG, and has therefore a slightly higher viscosity to start with (which means that it needs a bit higher temperature to print).
CPE+ is a PCTG , which has higher stability to degradation than PET(G) (cannot find a nice public reference to explain that now unfortunately). Also, PCTG likely absorbs less moisture than PETG, which means that there is less moisture (at similar conditions) in the material during printing.

So to sum up: the combination of moisture uptake, printing temperature and bond-type basically determine how much hydrolysis = stringing you get. Ideally you want to reduce each effect as much as possible (by drying your material before printing and selecting slightly lower printing temperatures if possible).
 

From my personal experience (not backed by data): PETG is most sensitive to stringing, then CPE and finally CPE+.

Hope this clarifies something!

[cheuschober 1]

Hi. I've included some photos.

The filament is *definitely* not moisture saturated as I dried with a print-dry and then printed from a drybox that's loaded with dessicant and has a short-run bowden straight to the ultimaker (with less than a couple hours out of the print-dry).

These are a mix of retraction, speed, and temperature towers. The PETG prints on a temp tower as low as 215C (which seems really low).

These results are similar for both the transparent (Matterhackers) and Ultimaker branded PET-G.

[geert_2 557]

On 11/15/2021 at 11:25 AM, TimonR said:

Hi @gr5,

In principle all condensation polymers (polyesters like PLA, PETG, CPE = branched PETG, CPE+ = PCTG, and polyamides/nylons) have this issue. These polymers are 'coupled' by connecting monomers, which releases water. The opposite (decoupling / hydrolysis) happens when you have moisture and high temperature.

See here a very nice explanation on viscosity decrease at higher temperatures (hydrolysis) of polyamides: http://extrusionist.com/presentations/prejudices about polyamide.html (from slide 20 onwards).

Now, it basically depends on the moisture uptake of these polymers how 'bad' the situation will be.
CPE is a branched PETG, and has therefore a slightly higher viscosity to start with (which means that it needs a bit higher temperature to print).
CPE+ is a PCTG , which has higher stability to degradation than PET(G) (cannot find a nice public reference to explain that now unfortunately). Also, PCTG likely absorbs less moisture than PETG, which means that there is less moisture (at similar conditions) in the material during printing.

So to sum up: the combination of moisture uptake, printing temperature and bond-type basically determine how much hydrolysis = stringing you get. Ideally you want to reduce each effect as much as possible (by drying your material before printing and selecting slightly lower printing temperatures if possible).
 

From my personal experience (not backed by data): PETG is most sensitive to stringing, then CPE and finally CPE+.

Hope this clarifies something!

 

Thanks for the explanation. Such deeper understanding is what is often lacking in manufacturer's manuals and tips on 3D-printing.

 

And in my view, personal experience is also "data". The first people to discover or to do something new, are collecting data by experience. So I think personal experience is equally valid, if not more, than citing references from others.

 

When molten, PET (or variations of it) stay way more rubbery and elastic than molten PLA. PLA gets a sort of yoghurt-like consistency, while PET is more like chewing gum: it strings more, but at the same time is more difficult to bridge gaps, because the strings tend to snap and fold back onto themself. Like a rubber band or chewing gum string that snaps.

 

Also, there are different sorts of strings:

- insect antennas: the nozzle leaks while traveling through air, and deposits that drop onto the next object it encounters. The next traveling move, it deposits its drop onto the previous drop, etc, causing the insect antenna. These are in your pictures too.

- big strings from the nozzle taking off from an object.

- thin hairs when material accumulates on the outside of the nozzle, and this gradually sags and sticks to other parts, pulling very thin hairs

 

Printing near the cooler edge of the temp-range, and slower, reduces these effects for me, but does not eliminate them. Because then the pressure in the nozzle is lower and the material is less liquid, so it leaks less.

 

[gr5 2,172]

The final print - the black one - has a cooling issue.  You need to print 2 of those parts at the same time so that one is cooling while it prints the other.  I have had very similar issues printing the two antennas on the Ultimaker robot head.  If I print 2 robots then it's all good.

[TimonR 21]

@cheuschober, thanks for adding some pictures.

Another tip for PETG: is you printcore clean on the outside? PETG has a tendency to stick to the outside surface of your nozzle and thereby draw strings as well - cleaning the outside (while hot) might help too.

Is your feeder clean? Might be that if you have some grinded material there it can affect the performance of your feeder?

What temperatures do you have for the Ultimaker PETG in your towers?
For AA0.4 with 0.15 layer height and engineering profiles I think we've seen good results between 230-240C printing temperatures. 

If you're unsure of the temperatures of your printer, you could try to follow the live data while printing test samples, see here how to do that: https://support.ultimaker.com/hc/en-us/articles/360021076960-Ultimaker-printer-APIs. 

- Retraction distances are typically 6.5, 8.0 or 10 mm for bowden tube printers. Possibly higher retraction distances could help you out?

Another crucial setting in Cura is the 'Max comb distance with no retract', which determines the maximum travel distance before it retracts. For materials like PETG you want to avoid unretracted travel moves, as they typically causes strings. 
We've set it at 8 mm, but you can set it also lower (to 4 mm) at the cost of more retracts and therefore more printing time.
Also, in Cura 4.12 there is bugfix that solves combing through skin (might be helpful for your specific PETG case: https://github.com/Ultimaker/Cura/releases).

- Object cooling:

Ultimaker printers typically have the biggest change in fan speed between 0 and 20%, anything over 20% doesn't really add anything in terms of cooling (except for more noise). 
Alternatively, you could use the 'enable bridge settings', to make an alternatively setting for enabling bridging if that's the only issue you're having.

Sorry for the not-so-concrete tips yet, but maybe these general ones can help you.

In general it should be possible to print PETG on most of our printers quite well (especially if you dry your materials well). 
If this doesn't work yet, feel free to share a part that you're trying to print without much success - maybe we can help.

[cheuschober 1]

11 hours ago, gr5 said:

The final print - the black one - has a cooling issue.  You need to print 2 of those parts at the same time so that one is cooling while it prints the other.  I have had very similar issues printing the two antennas on the Ultimaker robot head.  If I print 2 robots then it's all good.

Hi @gr5 that's actually a retraction test, so one part with two towers. The reason it's so blobby is because so much material oozes it ends up at a higher Z and the nozzle just starts pushing it around.

[cheuschober 1]

11 hours ago, TimonR said:

@cheuschober, thanks for adding some pictures.

Another tip for PETG: is you printcore clean on the outside? PETG has a tendency to stick to the outside surface of your nozzle and thereby draw strings as well - cleaning the outside (while hot) might help too.

Is your feeder clean? Might be that if you have some grinded material there it can affect the performance of your feeder?

What temperatures do you have for the Ultimaker PETG in your towers?
For AA0.4 with 0.15 layer height and engineering profiles I think we've seen good results between 230-240C printing temperatures. 

If you're unsure of the temperatures of your printer, you could try to follow the live data while printing test samples, see here how to do that: https://support.ultimaker.com/hc/en-us/articles/360021076960-Ultimaker-printer-APIs. 

- Retraction distances are typically 6.5, 8.0 or 10 mm for bowden tube printers. Possibly higher retraction distances could help you out?

Another crucial setting in Cura is the 'Max comb distance with no retract', which determines the maximum travel distance before it retracts. For materials like PETG you want to avoid unretracted travel moves, as they typically causes strings. 
We've set it at 8 mm, but you can set it also lower (to 4 mm) at the cost of more retracts and therefore more printing time.
Also, in Cura 4.12 there is bugfix that solves combing through skin (might be helpful for your specific PETG case: https://github.com/Ultimaker/Cura/releases).

- Object cooling:

Ultimaker printers typically have the biggest change in fan speed between 0 and 20%, anything over 20% doesn't really add anything in terms of cooling (except for more noise). 
Alternatively, you could use the 'enable bridge settings', to make an alternatively setting for enabling bridging if that's the only issue you're having.

Sorry for the not-so-concrete tips yet, but maybe these general ones can help you.

In general it should be possible to print PETG on most of our printers quite well (especially if you dry your materials well). 
If this doesn't work yet, feel free to share a part that you're trying to print without much success - maybe we can help.

 

Hi @TimonR -- glad to provide clarity.
 

1. Printcore is clean externally and I even replaced the silicone gasket.
2. Temperature is defaulted at 240C and I've tried calibrating with a temperature tower between 215 and 260 but there's so much stringing it's really hard to tell what the "right" temperature is.
3. I've tried retraction towers too and while I've had some good luck with stringing with very aggressive retraction, if you'll note the backside of the round-towers, there's some very obvious blobs that don't particularly diminish. That's the start/stop on the z-hop so, I assume, an artifact of more oozing?
4. Everything was printed with bridge-settings "On"

 

I can fire up another retract test if it's preferred. It tends to show some of the worst behavior.

 

For completeness, I'm using models from the Calibration Shapes plugin (https://github.com/5axes/Calibration-Shapes)

[cheuschober 1]

Tried it again. Still oozed out a lot at the start and ended up getting caught on the nozzle, again, and pushed around.

Retract Tower.3mf

Edited November 17, 2021 by cheuschober

[gr5 2,172]

So is that a retraction test?  Is this telling you how much to retract?  Does it retract more and more as you go up the towers?  If so, what is that perfect retraction distance?

 

If not, if this is a print with normal settings, then I'd say it's too hot near the bottom (where you are closer to the heated bed and also where the fan might not be up to full speed - typically 5 layers to get to full fan speed).

 

[cheuschober 1]

@gr5 This isn't a retraction tower so it's not changing settings; it's standard ultimaker PET-G settings all the way with the one exception of changing brim for a skirt.

Oozing happens at a lot of spots. The print bed is often full of little filament dots from the automatic leveling. Then, there's a drag line from the bed front-center to where the print is starting while the filament head is heating after automatic leveling. I'd also offer that the quality further up still isn't great, some stringing and blobs at layer-starts is something I've observed quite a bit.

I think your theory might be correct regarding it being hot but I'm just a bit suspicious of something more given that I'm using stock settings and filament which I would expect to be well-tuned lest the forums be filled with issues of poor print quality.

I can try turning down the print bed settings or having the fan start aggressively.

[gr5 2,172]

Well your part is smaller than average.  When I have to print two tiny things on each layer (such as this tower) I usually get even worse quality.  So what I do is print two of the parts (so you would have 4 towers) and this makes a massive difference.  It gives the part just another second or two to cool.  Those profiles are limited and tend to work for average parts.  There isn't a profile for "printing a part smaller than a pea in diameter".  For such a small part you need to tweak the existing profiles.

 

I wouldn't mess with bed temp but I'd get the fan started a little sooner I think?  Maybe.  I'd check the settings to see what fan speed it is at for each layer.  Typically I think it has changed - it once was 5mm before the fan was at full speed and I think more recently it must be closer to 1 or 2mm?  Or 5 layers?  I just don't know.

 

The other issue is that the percentage fan speeds don't make sense at least for my UM3 and S5.  For those printers the fan is typically at full power when it is around 10 or 20% (depending on the printer - they are different).  And there is no increase in speed going up from there.  So I consider 3% as a good low speed, 6% medium, and 15% pretty much full out fan.  But I don't know if my printers are unusual or not.  But knowing this and looking at the profiles out there, something is wrong with the fan speeds of many profiles.  At least for my S5 and my UM3.

[TimonR 21]

Was in the office today, so quickly printed your @cheuschober samples as well with Black Ultimaker PETG on S5.

 

On the left you see the standard 0.15 Layer height PETG profile of Ultimaker PETG. Indeed, I also see a blob, but only on the bottom few layers of the tapered pillar only  (not on the straight pillar).

 

If I run your 3mf, I get the printed part on the right, where in my case the blob is absent.

For the rest, stringing is fairly minimal (but both have a few strings halfway).

 

I'm not sure exactly if this helps you, but it seems to be an issue in your printer?
Hope this helps?

 

I do agree with @gr5 in the post above here, that this is really a small part to do 'optimization' on. Typically we use parts that are at least 2 to 10x bigger to optimize for stringing, so we don't have these very small details.

Possibly such a part would benefit from an AA0.25 - if you actually have such small dimensions?