TPU Filament Getting Jammed

[bostonman]

Has anyone experienced TPU filament getting jammed?

Months ago when I figure purchased this roll, I printed some test stuff at 0.3mm layer height (0.35mm first layer height).

Today I tried printing something at 0.15mm layer height, 210 degrees C at first, and then 220 degrees C after. After three or four attempts, the TPU was coming out of the side of the nozzle housing (Anet A8).

Initially I thought 210 degrees C was too cold and causing it to seize in the nozzle, so I increased it to 220. After I thought the roll wasn't spinning easy enough and causing too much resistance, so I unrolled slack.

My guess is the 0.15mm layer height is causing the filament to bunch up inside the nozzle since I didn't seem to have issues with printing a 0.3mm.

Has anyone experienced this issue and how should I resolve it?

[Whales]

Yes.  On my cheap printer TPU filament starts coming out of the extruder sideways if too much extrusion force is needed to get it through the nozzle.

A better extruder design is the most effective solution.  My design is borderline, it works IF I put the temperature up a bit higher, print slower and make sure the first layer has a big gap between nozzle and bed.

[CatalinaWOW]

I have also encountered TPU jamming under unusual circumstances. A glitch on the power line caused the print to stop.  With the fans not running the heat soaked back up the filament path and softened the filament.  I did not realize the heat soak issue a restarted the print.  The filament retracted and kinked inside the direct drive extruder.  Any similar issues might cause a similar issue.

[MarkF]

I've not printed TPU.  But, I've seen where the gap between the extruder gear and the hotend is too big and the flexible TPU squishes out.  There are various ways to try to fix this.  Take a look at this video to see what I'm talking about.

https://youtu.be/6kZkxkSRndM?t=170

Every brand of TPU will act differently.  What brand are you using?
For temperatures, I usually start with values in the middle of the manufacture's recommendations.

Do you know what the hardness is of the TPU that you are using?
See...

https://youtu.be/iWZw7RO2Sks?t=168

[bostonman]

MarkF,

Just an FYI, I was trying to print the Agilent knob you created for me. My goal was to see if it would be hard enough that it can't be squeezed, but have the subtle rubber grip as the original.

I'll check the filament brand and stuff later.

[radar_macgyver]

TPU is prone to moisture ingress. If your spool was not humidity controlled, the moisture could boil off inside the extruder and result in jams.

[MarkF]

Just an FYI, I was trying to print the Agilent knob you created for me. My goal was to see if it would be hard enough that it can't be squeezed, but have the subtle rubber grip as the original.

I'll check the filament brand and stuff later.

That's what I suspected.
Overall, I believe TPU will be too soft and flexible for a knob.
That's why I asked about the TPU hardness you're using.  A different brand may be desirable.

In the first video, he made a print at 235^oC.
The print turned out to be harder and less shiny.  Both desired characteristics in this case.
A high percentage of infill should make it harder also.
Try a print 5^o or 10^o hotter to see what you get.

If you look close at the bottom edge of your knob, it looks like it's made in two piece.
A hard inner core, wrapped with a soft cover. 
However, printing a hollow soft outer cover for a knob would be problematic.

[bostonman]

This is the stuff I bought (description directly form Amazon):

GIANTARM TPU Filament, 95A TPU Filament 1.75mm, Dimensional Accuracy +/- 0.03, Flexible Soft 3D Printer Filament 1kg Spool(2.2 lbs), Vacuum Packaging (Clear)

[MarkF]

The 'shore hardness' of 95A is what I was curious about.
It will be interesting to see how you make out.

Amazon has TPU with a 98A shore hardness if your knob ends up being too soft.

[jm_araujo]

Not knowing your slicer, are you using a TPU profile and/or did you change the speeds?
I don't have much experience with TPU, but when I tried it I found that in my old printer (bowden extruder noname chinese) I had to slow everything down by at least 50%, or else something will fail.

[bostonman]

I don't have much experience with TPU, but when I tried it I found that in my old printer (bowden extruder noname chinese) I had to slow everything down by at least 50%, or else something will fail.

Maybe that's another problem. I never play with any of the settings, usually I just change infill, temp, bed temp (220 for extruder and 50 on the bed for TPU), and layer heights. Edit my slicing software doesn't have a TPU setting or PLA setting.

Since this is a small knob, maybe the filament is being fed too fast and jamming where as before the gasket I downloaded to do a test print was larger, so the head was moving faster.

Due to reducing layer height to 0.15 and the knob being small, it's possible feed speed is the issue. I don't know for sure, but it seems the failure occurs around the same height in all three cases.

[MarkF]

From the Amazon page for the GIANTARM TPU:

• Temperature:      195 - 230^oC
• Bed Temperature:  40 ~ 60^oC
• Print Speed:      30 ~ 45mm/s
• Layer Height:     >= 0.25mm
• Nozzle:           >= 0.5mm

You might want to try:

  230^oC temp, 50^oC bed, 30mm/s speed, and 0.20mm layer (or 0.16mm)
  At least knock off 10mm/s from each speed value in your slicer

[abeyer]

Overall, I believe TPU will be too soft and flexible for a knob.
That's why I asked about the TPU hardness you're using.  A different brand may be desirable.

I haven't had a chance to try it yet... but I have access to a prusaxl with multiple print heads now, and want to experiment doing some multi-material knobs and buttons with both a rigid structure in petg/pla/etc and a tpu surface as a monolithic print. Seems like that could be the best of both worlds and give a result similar to the overmolding processes often used to produce that type of thing commercially.

[bostonman]

Nozzle and bed temperatures were basically already set to those values because I saw that on Amazon and/or the spool tag. Initially I may have accidentally ran at 200 degrees C the first time it jammed, but then changed it to 220 and then 225 based off your advice (although you said 230). Bed temperature has been at 50 too and the nozzle setting is a default 0.4mm. I'm uncertain if my nozzle is 0.4 or 0.5, but I've never tinkered with this setting since I never had clogging issues until now.

Also, I had tinkered with layer height by decreasing it to 0.15, but changed it to to your recommended 0.25mm.

Unfortunately it jammed again with basically all the settings suggested, this time at maybe a layer or two higher than before, but still jammed.

When I pull out the filament, after it unwinds from looping to the outside, it makes sort of a suction sound leading me to believe it's feeding too slow and getting bunched up inside the nozzle.

Ignoring that I doubt printing a knob with TPU will be ideal (I think you're correct MarkF, it will be too soft), the part may be too small so the retracting is causing the end to curl and/or it's getting curled due to sitting in the nozzle for long periods while the head prints the small sections of the knob.

Attached are the default values of my slicer. It's unclear which is the "print speed", but maybe the one I tweaked to 30mm/s was incorrect.

MarkF, on a side note, I didn't realize TPU came in different shore ratings. Buying the TPU filament was a spur of the moment decision to tinker with different material, so I didn't investigate at all. It does seem the knob either has a rubber overlay or was dipped in something (good eye). After looking closer, the main section is hard plastic and then I see a small line indicating different material. So a fully printed TPU knob may not be similar unless, as you suggested, I go with a 98A, but then I'm diving into buying approx. $20 spools to experiment with printing a knob.

[Ian.M]

How close is your extruder's throat tube to the extruder gear and pinch bearing?  For flexible filaments, there should only be just enough clearance for the gear to turn freely and the pinch bearing not to contact the throat tube with no filament present.  See:
https://youtu.be/B71GGLBrQDU?t=82

[MarkF]

First, have you done anything to eliminate the gap between the extruder gear and the hotend as described in that first video I posted?  This insures that the soft TPU is pushed into the hotend and doesn't bend away.
Before and after screenshots from video:



On the speeds, there are more than just one speed.  A slicer normally has multiple speeds depending on what it is doing at a particular time.  For instance, one for traveling (just moving the head), one for printing a perimeter, another for printing in-fill, etc.  Here are my PrusaSlicer settings for normal PLA:



In this case for a knob I printed for my model railroad throttle.
You are already printing much faster than I print for PLA.  I can't push PLA to those speeds with my Creality CR-10.
I would suggest knocking down all of the speeds in the 'Speed for print moves' section. 
Start by limiting the speeds to 30mm/s. 

Most printers come with a 0.4mm nozzle.  So, unless you changed it?

[jm_araujo]

Try halving all the speeds. In your settings they are "Perimeters" (set to 30), "Infill" (set to 40) and  "Bridges" (set to 30).
You could leave as is "Supports" (which you probably are not using), and "First Layer speed" as you said you didn't have problems in the first layers.

[MarkF]

And, I suggested pushing the high end of the temperature range because the video said the higher print temperature made the print harder.  So, high temperature and a lot of infill to make you knob harder.  You are NOT looking for the typical soft flexible TPU print use case here.

[bostonman]

First, have you done anything to eliminate the gap between the extruder gear and the hotend as described in that first video I posted?

I first wanted to rule out print settings before dismantling and making mechanical adjustments. Unfortunately those settings didn't work either, so it's off to figure out the gap.

MarkF,

This time I got a bit more of the knob printed; not much more but more. Using 230 degrees C either made it stiffer or due to getting more physically printed. The shaft hole doesn't seem to have the same resolution as PLA, but, due to it stretching, might fit. I was thinking, would it be possible to make the inside more hollow like the original knob and then I can fill it with two-part epoxy?

This way the outside will be rubber, but it can't be squeezed due to the hard epoxy.

Just a thought. You've already done quite a bit helping with two knobs, so I'm grateful to have the solution in place (i.e. you helping with the design and living in a world with 3D printers).

[MarkF]

This time I got a bit more of the knob printed; not much more but more. Using 230 degrees C either made it stiffer or due to getting more physically printed. The shaft hole doesn't seem to have the same resolution as PLA, but, due to it stretching, might fit. I was thinking, would it be possible to make the inside more hollow like the original knob and then I can fill it with two-part epoxy?

Not sure I understand what you want.
In your slicer (at least in the PrusaSlicer and Cura slicers) you can use a 'grid' infill and set the number of bottom layers to zero.



The higher hotend temperature will make it easier to push the filament threw.

Don't forget to go to the Thingiverse link in the video and printout that filler part before tearing apart your printer.   

[bostonman]

The STL you provided shows a solid (see attached) inside in the 3D view.

I set infill to 80% to make it harder, but, since the 3D model view showed a solid, I thought you just made it solid all the way through to save time and didn't pay attention to the possibility of it being more hollow.

My slicer software gives the option of how many top and bottom layers to make solid. I set the bottom to 0, and, if it doesn't cause the software to crash, allows me to do as you stated.

If I print it as you suggested, this may be a pretty cool knob providing I can get some two-part epoxy inside to stiffen it. Maybe I'll invest in a spool of gray TPU, and, with some tweaking, may not even notice the knob isn't original.

Update: hopefully I'm not too late updating my post. I'll gladly print that filler part, but will it melt/bend/warp sitting in the hot end area that's 230 degrees C?

[MarkF]

FreeCAD and other 3D modelling software produce a solid object unless you've carved out interior holes.
Your slicer settings determines how the inside is printed.  (i.e. the interior pattern and percent of fill)
Through the slicer you can make it hollow, solid, or something in between.


I have a bowden tube setup on my CR-10.

If I understand the Anet, you should have the extruder drive gear -> heatbreak -> hotend (heating element) -> nozzle.
That part goes in the drive gear mechanism, before the heatbreak.  It should never get hot unless your fan fails.

[MarkF]

Here is the modification on a bowden setup:



Notice how the point of the modified part (highlighted in red) goes right in between the gear and the idler preventing the TPU from escaping from being pushed into the hotend.  The Anet mod only has the triangle wedge.  The filament is then fed into the hotend and nozzle.  It should never get hot.

[CarolinaCook]

Oh man, I’ve definitely had issues with TPU jamming before, especially on an Anet A8. It’s super frustrating. The lower layer height could be causing the problem since TPU is so flexible. When I had similar issues, it turned out that the filament was getting bunched up in the nozzle, particularly when trying to print finer details.

Increasing the temperature might help a bit, but I also found that printing slower helped a lot. TPU can be tricky, but once you find the sweet spot, it’s totally worth it. Have you tried tweaking the print speed or maybe checking the filament path for any snags? Sometimes even small adjustments can make a big difference. Good luck

[bostonman]

Increasing the temperature might help a bit, but I also found that printing slower helped a lot. TPU can be tricky, but once you find the sweet spot, it’s totally worth it. Have you tried tweaking the print speed or maybe checking the filament path for any snags? Sometimes even small adjustments can make a big difference. Good luck

Thanks for the feedback.

Others have provided similar feedback and the current most likely solution is to dismantle the head, check the spacing, and install a 3D printed TPU guide. This suggestion makes perfect sense because if the feed gear (?) is too far from the nozzle, the TPU will begin curling. Reducing the space seems the most likely solution.

Unfortunately I've been over my head in projects and haven't had a chance to return to the printer. Long story short, this issue began because I'm printing a replacement knob with TPU for a recently repaired oscilloscope. Due to some issues, I've been elbows deep with trying to get the scope reassembled; thus delaying the need for the knob. Thankfully the scope is fully reassembled, needs to do some self calibration and tests, but soon I can divert my attention to the printer.

Also, the 3D TPU Guide part is suggested to print in ABS. I have a roll, however, I haven't opened it yet nor have experience with ABS.