Printing TPU tips?

[Jerry-rigged]

Well, my Overture print is done, it looks a lot like that last Ninjaflex print. A bit more stringing, and the corners still lifted.

The Overture came with a build surface, I'll try that before I try a brim.

Looking at your model, yeah, I like the lower zip-tie chamfer closer to 45deg. But why did you add the chamfer across the bottom?

Latest try -

20200930_140100.jpg

• Jerry-rigged
• Sep 30, 2020

 

[suchtragedy]

Yeah I'm gonna vote that your machine needs proper e-steps calibration. 95A hardness tpu is difficult to get wrong so if it's that bad my vote is you're not extruding enough material because the calibration is off.

I would give up on printing this and work on some test calibration type prints first until you get it right and then go back to this case once dialed in.

 

[Jerry-rigged]

@lycin8008 - Printing your model now. Same settings, so I am guessing it will still string (but I will live with that for now). On the Overture build plate this time. Was interesting leveling it on that stuff, I didn't realize I was relying on the mirror to see the bottom of the brim (I use paper to get close, then level while printing the brim).

Yeah I'm gonna vote that your machine needs proper e-steps calibration. 95A hardness tpu is difficult to get wrong so if it's that bad my vote is you're not extruding enough material because the calibration is off.

I would give up on printing this and work on some test calibration type prints first until you get it right and then go back to this case once dialed in.

Good idea. If I can just get the corners to stick, I think I will get a workable, but ugly print. Then I will dive into calibration prints. Honestly, I have not done much of that, because with PLA, the only issue I ever have is sometimes (1/5 or 1/10) it will not stick to the bed. I mean, I have a small fleet of Benchies, but they always print fine... LOL

 

[lycin8008]

Looking at your model, yeah, I like the lower zip-tie chamfer closer to 45deg. But why did you add the chamfer across the bottom?

If you look at your original model, the edges curved up at a very slight angle at first, ~5-10 deg, that is essentially starting the "lifting" problem. Prints, especially with flexibles prefer an angle closer to 90 deg, but you can get away easily with 45. When printing in general, as the hot end goes over the layer below, it kind of pulls on the layer. With the material being flexible, it allows that layer to get pulled more, then cools not where it was supposed to be. The more material that is directly below it, the more downward support it has to stay put.

 

[Jerry-rigged]

You know, I forgot the bottom had the fillet. Dang Tinkercad won't let you pick the edge to fillet, it is all or nothing with that app. Quick to knock something simple out, but very limiting. I need to DL Fusion. Funny, CAD is my day job (when I have a job). Learning Revit now to hopefully help get a job, so I have a student Autodesk license for a year. Once I get thru the Revit class, I need to get Fusion running.

Edit-
I hate it when that happens. Guess the bed was not as level as I thought. LOL

 

[lycin8008]

I hate it when that happens. Guess the bed was not as level as I thought. LOL

I had the same issue with those free surfaces and brand new BuildTak. You have to clean them with alcohol to get any manufacturing residue off. I also use a little glue stick with it as well.

 

[Jerry-rigged]

If I still need glue stick or spray, then what is the point? LOL Printing on the glass has always (mostly always) worked, but guess I just wanted to have an excuse to try something new...

 

[fpvmiller]

Here's a tip for leveling: Don't use a sheet of paper.

I think that a sheet of paper or any distance between the nozzle and bed is wrong.

When leveling get your nozzle to touch the build plate. I use probes with mesh leveling and my z-offset is set to the nozzle touching.

Why have the nozzle touch the bed versus using paper? When you home your z it homes to your 0 position. When you do a g0 z0 your nozzle should be touching the bed at 0. This way when your first layer say at 0.2 layer height starts printing you are actually printing at 0.2 and not 0.3 or whatever.

My first layers are always spot on. I don't use glass or painters tape. I use buildtak and I don't use glue or hairspray either. I have to take my prints off when the bed is still hot because it sticks that well.

 

[lycin8008]

I use the large sheet of BuildTak on my 310mm^3 printer since i print on a sheet of spring steel, glass isn't flexible, esp. on large parts, and i print in PETG a lot too. I use a pane of glass between the heater and the flexible surface since it had lower deflection when heat cycling.

Here's a tip for leveling: Don't use a sheet of paper.

I think that a sheet of paper or any distance between the nozzle and bed is wrong.

When leveling get your nozzle to touch the build plate. I use probes with mesh leveling and my z-offset is set to the nozzle touching.

Why have the nozzle touch the bed versus using paper? When you home your z it homes to your 0 position. When you do a g0 z0 your nozzle should be touching the bed at 0. This way when your first layer say at 0.2 layer height starts printing you are actually printing at 0.2 and not 0.3 or whatever.

My first layers are always spot on. I don't use glass or painters tape. I use buildtak and I don't use glue or hairspray either. I have to take my prints off when the bed is still hot because it sticks that well.

The paper trick is good if you don't have auto mesh leveling, especially when using it on the middle and four quadrants of the build surface. As for the accuracy of the first layer, this is FDM 3D printing, not machining, so two tenths of a millimeter more in the Z-axis should not affect your part so much. Ghosting/ringing would affect your dimensional accuracy too, unless you are one of the guys chasing the perfect print. Even my Stratasys uPrint doesn't print perfect parts. I made a nearly perfect 3DBenchy out of TPU on a hobby level printer that was better that the uPrint's benchy with its native ABS with dissolvable support material (cant turn supports off).
I guess it's the hobbyist/operator's call on what is good enough.

 

[Jerry-rigged]

My printer is a regular old cr10. No Auto bed level, no mesh level, nothing fancy. I just use the sheet of paper to make sure the nozzle will not drag the glass. When I slice a print, I always add a 10 Loop Brim. While they brim is printing, I will I eyeball the level and adjust as needed.
Here is the first layer on the print that is going right now. The Bram shows color variation, where I was adjusting it every few laps...
I think I have a bubble of air under that build surface. In the lettering on the right, you can see where the brim thins out very much. The rest of the bran is very consistent.

 

[Jerry-rigged]

This last print look useable. Still a good amount of stringing, guess I lower the temps to control that? Overhangs were still not pretty, but better than before. With a bit of clean up, it would work, though.

BUT - the build surface - it won't let go! I broke the model trying to get it off - a wall split.... LOL Ended up pulling the whole sticker off. Model is still stuck.

20201001_061946.jpg

• Jerry-rigged
• Oct 1, 2020

Edit - got it un-stuck. Ripped a few more walls. Spent a few minutes cleaning up the strings. With a bit of superglue, I think this should be a workable model...

Thanks for everyone's help. Off to Thingiverse to find some calibration STLs.

 

[jondilly1974]

Check out Teaching Techs videos with a flat X

 

[lycin8008]

This last print look useable. Still a good amount of stringing, guess I lower the temps to control that? Overhangs were still not pretty, but better than before. With a bit of clean up, it would work, though.

BUT - the build surface - it won't let go! I broke the model trying to get it off - a wall split.... LOL Ended up pulling the whole sticker off. Model is still stuck.

Edit - got it un-stuck. Ripped a few more walls. Spent a few minutes cleaning up the strings. With a bit of superglue, I think this should be a workable model...

Thanks for everyone's help. Off to Thingiverse to find some calibration STLs.

Layer lines look uneven. When was the last time you oiled the V-wheel bearings? or adjusted the eccentric nuts?
Flat spots on the wheels and loose sleds can give some interesting artifacts.
But I agree with your earlier sentiment, there is a bit of calibration you have to do before you can get excellent prints.

 

[Jerry-rigged]

Yeah, not much maintenance has been done on my CR10. But then, I have not used it as much as I thought I would. Most of my printing has been "that looks cool" knick-knacks that I didn't stress over when they didn't look perfect, or quad frames. Again, I didn't care about layer lines...

But, I got lots of time now to play with calibration and tuning.

 

[suchtragedy]

Did a few quick edits https://www.tinkercad.com/things/1GT1y0eheN4

@Jerry-rigged I think I will try printing this version by @lycin8008 today and once I get it good using very similar material if not 100% the same I will post my gcode file here, then have you try printing my gcode file without doing any changes and then see how it turns out. The purpose of this is to eliminate the variable if it's slicer related or if it's machine calibration/hardware/maintenance related. If the print comes out clean then it's slicer, if it's still junk it's the machine/something physical somehow.

 

[lycin8008]

Just printed it. 1.5 hrs. 0.26 mm layer height, 0.6 mm nozzle at 220°C and no part cooling.
40mm/s infill and inner walls, 20mm/s outer wall, 50mm/s bridges.
4 shells (not really needed), 3 top and bottom layers with 1 surface skin layer and 30% infill (cubic pattern).
Using Sainsmart black TPU.

 

[Jerry-rigged]

Looks pretty clean!

 

[Camaroboy383]

Hate to say this, but to me it looks like a classic nozzle clog... the byproduct of pla's is most likely a nasty mess in there.. you will never get the extrusion correct if that gummy nasty junk is hanging out in the throat of the hotend..

How long since you serviced it?

 

[lycin8008]

Not sure who else here recommends it, but I have had good results from lubricating the hotend, especially on a new throat/heatbrake tube, to help reduce the sticking and buildup on the inside. I know PETG and TPU can get pretty gummy on an older hotend.

 

[Jerry-rigged]

Hate to say this, but to me it looks like a classic nozzle clog... the byproduct of pla's is most likely a nasty mess in there.. you will never get the extrusion correct if that gummy nasty junk is hanging out in the throat of the hotend..

How long since you serviced it?

I swapped in a new nozzle not long ago, and I don't think this one has had any PLA thru it. Mostly TPU and a bit of PETG.



So now I am trying to print one of those temperature towers. Never done one. Can the Creality printers (mine is a CR-10) use a script to change the temp, or do I need to mine the gcode to find the layer changes and insert a M104 (temp code) manually?