Anyone having Success Printing TPU with Dragon HF?

[VoronManiac]

On a Voron 2.4 SB, I'm printing some large skinny panel frame pieces that use ~12M of filament  and are mostly perimeters with a small amount of infill.  Flow is limited to 3mm^3/sec. PA=0.15. I typically see 3 or 4 surface defects in the 4 hour print. I captured one of the defects forming in the pic below a few layers after it occurred.  It appears the extrusion stopped for a couple of mm and then a blob was ejected.  The DHF has a barrel diameter of 1.90mm (according to their specs and also measured) which in on the small side compared to most other extruder barrels 1.95 - 2.00mm.  Most TPU filaments I have seen tend to have some wiggle compared to PLA & ABS.  When I take a length of Zyltech filament with an OD of 1.78 and try pushing it through a brand new DHF, it does slide, but I'd describe the fit as snug (pic below).  Just because the cross section "diameter" is 1.78, vendors do not spec concentricity.  Additionally the filament picks up indents when it passes through the drive gears.  Too little drive gear tension and the filament slips.  Too much and it ends up with large indents.  When I switched out the DHF for a E3D V6 (Formbot clone) which has a 2mm barrel diameter, pretty much all the print surface defects seem to disappear.

I had been pre-screening rolls of filament by running then through the new DFH extruder looking for high spots and respooling.  This seems extreme.  But it did catch what looked like a splice in the middle of a roll of Yousu TPU.  I had two rolls of Paramount TPU that were wildly out of spec at 1.92mm that needless to say were causing hard jams.  Customer support was  looking into it but haven't hear anything back for a few months now.  I have observed some tight sections of Zytech TPU that would probably cause issues.  These sections were within spec, but might have had concentricity or wiggle issues.  I have a roll of Matterhackers that I should try screening.  As tricky as it is to print TPU, I do feel for the manufacturing engineers tying to make this stuff because it is so flexible.

I'd like to hear about success stories brands, tension and special print settings.  I am trying to favor domestic suppliers with prices less than $30/kilo but at the end of the day, take what ever works.  For now reverting back the the E3D seems like a good fix.

 

[mheath82]

I print a lot of large parts in TPU 95A on my SB w/ Dragon high flow, but I have to slow it down to max of 60mm/s and 10 MM3 or it goes terribly wrong. I also had to use the internal tension screw to apply less pressure on the filament, about 1.5 turns if I recall. I'm just using a generic brand 95A.

I'm still tweaking but this is what I've been able to print reliably with.

Edited March 17, 2023 by mheath82

[SVDB]

I print Filright TPU 98A slow @ 70mm/s using a Bondtech LGX and a high flow Dragon HE and a .6 nozzle.

Since I mainly print gaskets and seals that are going to be sandwiched between parts I never really cared for the surface finish quality (which is still decent but not top notch like for a phone case), so I didn't put a lot of effort in the profile, but it always prints on the first try and I never had any blobs or failures.

 

[2tekno]

On 3/17/2023 at 12:40 PM, mheath82 said:

I print a lot of large parts in TPU 95A on my SB w/ Dragon high flow, but I have to slow it down to max of 60mm/s and 10 MM3 or it goes terribly wrong. I also had to use the internal tension screw to apply less pressure on the filament, about 1.5 turns if I recall. I'm just using a generic brand 95A.

I'm still tweaking but this is what I've been able to print reliably with.

This: "to apply less pressure on the filament.." helped me to successfully print in TPU - thank you for that advice! I'm just wondering what spool holder do you use for your setup? 

[VoronManiac]

Agreed. Tension setting can be an issue.  Needs to be much less than ABS.

I'm getting good prints now with ZYLtech TPU with the E3D V6.  Dragon HF is a lost cause.  I had been using the standard Voron spool holder which works well enough, but it is exposed to the air allowing the TPU to absorb water even with 15% (winter time) relative humidity - not dry enough. I started pre-drying the filament in a food dehydrator with pretty uniform air flow 150F which was getting into the 5% RH range.  I then started using a home made drybox with "Wisesorb 7.5 LBS Activated Alumina, Alumina Desiccant, Reusable Desiccant" from Amazon.  The other piece of magic was to increase retract to 2.5mm.  The blobbing was probably caused by  a combination of stringing and exploding micro steam bubbles causing blobs to build up on the nozzle.  Prints now have absolutely -0- stringing.  Going forward, I always dry filament TPU or ABS if I care about the print quality.

Lately I switched to the EIBOS single roll filament dryer.  It has a LOT of issues. Not sure I recommend it, but it does properly vent in ambient air.  Swapped out the Humidity sensor for combo temp/humidity.  The different types of filament settings on it are a joke because of the uneven temperature/air distribution.  They didn't put temperature settings in degrees on it because the temperature varies wildly throughout the dryer.  Performance is extremely dependent on width and size of the center hole of the spool which controls airflow. You can use online humidity tables to calculate how much the R.H. drops at a give temperature in the box given ambient.  The cheapy R.H. senors don't go below 10%, but knowing the temperature gives the the R.H. if you know your ambient.  I printed a custom door extender (like others on line)  to partly to accommodate wider spools, but also to improve air flow.  The rubber band/o-ring that comes with the box to hold the spools on is yet another joke - printed a retainer clip replacement.  It is possible to deform ABS is you crank the temp up all the way AND if the airflow is restricted because of the spool.  For TPU with good air flow, I crank the temp up to 125F which is still not as good as a food dehydrator but seems good enough.  The humidity sensor does not indicate humidity of your filament roll, it is just the R.H. of the air drying the filament which is a function of how hot everything is.

[claudermilk]

I finally broke down and got the fancy Sunlu one. It's worked well for me; my nasty, soaking-wet, spitting TPU and Nylon both got dried and printed decently after a couple runs through it. I even had a problem spool of PLA that I ran the PLA cycle & it printed great again. It's a bit spendy, but I'm happy with the purchase.