Run Out of "Knobs" to Turn

[Gwhite]

I'm printing some eSun PLA+ "Fire Engine Red" parts, and I'm having a hard time getting the print quality dialed in. I have done a fairly systematic sweep of a wide range of variables: temperature, retraction, extra restart distance, coasting, printing in-to-out & out-to-in.

The part is small, with a conical section and an area off to the side that uses support:

The problem is that I get a groove where the extruder leaves & returns from printing the support section:

The best I've managed is printing at 215C, with 0.5mm retraction, +0.15mm extra restart distance, no coasting and printing in-to-out. I can't tell if the groove is the result of under-extruding before it goes off to print the support, or when it comes back (or both...)

Is there a parameter I've forgotten about I can play with? Other ideas?

Thanks!

[ednisley]

Other ideas?

You're trying to get a specific surface finish out of molten goo: fighting the process can't end well. Seen from here, the surface finish is Just Fine™ and will never get better … you can only make it different.

Instead, work with the process: if finish is important, optimize the part and its layout with the process in mind.

Quick fixes:

Put the cone tip in the air, with a millimeter of support under the arm. (Works with hollow cone)

Put the arm straight up, with small support under the cone tip. (Assumes the cone is solid)

Better:

Cut off the arm (in the model!), put a slot for a locating pin in each section, print both with no support (cone big end down, bar standing on end), epoxy together around pin. Because you care about surface continuity, print one part at a time or print sequentially with manual layout and careful previewing.

Prediction:

The infill pattern varies as the cross-section of the cone and bar change, so your next issue will be small surface discontinuities on adjacent layers with different patterns.

[Gwhite]

The model is chopped off to speed up test prints. It's vaguely symmetric, and if I invert it, the section above where stopped the test prints will just do more of the same.

The overall print is usable, but the groove left in the part will collect grime and be hard to clean. I'm mostly trying to learn the limitations of the process. I suspect I would have the same issue if I printed two adjacent cylinders (an experiment I will try shortly). The print process works great for a single contiguous part, the issue is with starting & stopping the flow of filament when the extruder takes off & returns from a 2nd isolated section of the model.

I thought about it last night, and I have not experimented with is various speed settings. I'm guessing that zipping back & forth between the two sections isn't allowing the extruder to do its job properly, so I will try slowing the XY speed down (currently at 18000 mm/min) and see what happens.

[ednisley]

the groove left in the part will collect grime and be hard to clean

C'mon, the entire printed object consists of ridged edges and is porous to fluids, particularly through the top and bottom surafaces. If you want "easy to clean", coat it with enough epoxy to smooth and seal the surface, making any minor changes in the surface texture irrelevant.

Not, of course, that you'll find (cheap) epoxy to be impregnable. I covered an aluminum cheese slicer with epoxy a year ago and it's been quietly corroding right though the sealed surface:
https://softsolder.com/2018/10/01/month ... ontinuing/

starting & stopping the flow of filament when the extruder takes off & returns

Which is why I suggested printing the sections separately. The surface finish of the molten goo depends critically on hot end pressure, so any change makes a difference. Minimizing those changes is part of working with the process, not against it.

[Gwhite]

The parts are small, and I need to make dozens of them. Messing with two piece designs isn't worth the bother. Same for coating them. They will be fine as is, but if I can make them less prone to collection grunge with a simple tweak to the process, I'm curious to find out what I can do.

I've done some speed tests on cylinders. On a pair of 10mm diameter cylinders, I get a fairly clean small groove where it goes back & forth between them. Halving the default XY speed made only a tiny difference (if any). I noticed on the original part that the groove gets wider as the diameter of the cone increases. I printed one 10mm & one 20mm cylinder at 1/4 speed, and the groove in the 20mm was only a tiny bit bigger (if that), so it doesn't seem to scale directly. XY speed seems to have no noticeable effect.

[airscapes]

the grove is the Z axis change to the next layer. I have never gotten rid of it no matter what I have done. Orient your part so the the small part sticking off with the support is facing the back of the printer when you are facing the printer and the spool is on your left.
Then on the layer tab on the right side choose start point closest to specific location and enter X 100 Y 250.
With the part in the center line of the printer the Z change will be at the back end. Not s cure but sometimes looks better.
Let me know if you get it to go away and how you did it. Note if you have not calibrated your extrusion multiplier with a single wall square and digital caliper do so, as when properly set the grove is reduced. Also on the advance tab (assuming S3D) make sure only retract when crossing open space is checked.
Also, when printing tiny stuff the .25mm nozzle will improve your outcome utilizing an extrude width of .25, printing at 25-30mmS and a layer height of .14 (that worked best for the miniatures I printed)

[ednisley]

less prone to … grunge

I'd expect cleaning a barely perceptible groove will produce better results than digging grunge out of the corners around the arm (and whatever other features you have). If cleaning the part is such a problem, throw it out when it gets sufficiently dirty.

After all, we're living in the future and can make whatever we need, whenever we need it. [grin]

XY speed seems to have no noticeable effect

The hot end is basically a bucket of molten goo with a hole in the bottom and the longer it's unpressurized, the more it leaks. Consistent thread extrusion requires consistent hot end pressure, but each retraction cycle moves a fixed filament length and doesn't control the hot end pressure.

Put 'em together: the less time the hot end spends not extruding, the better the outcome.

[Gwhite]

the grove is the Z axis change to the next layer. ...

Actually, it isn't just the layer change. The layer change groove/line I'm used to, but there is a much larger defect when the layer change is combined with going back & forth to the support section:

As Ed pointed out, the pause between printing the regular & support section is longer than just going from layer to layer, which could be making things worse. However, it doesn't appear to be affected by the XY speed. If Ed's theory is correct, the higher XY speed should help, because the extruder has less time when it isn't extruding. Also, the distance between the regular & support sections is least where the defect is worst, which also implies that the problem isn't with the idle time between sections.

I've also tried slowing down the overall printing speed by a factor of two, with no noticeable improvement. I'm about to chalk it up to "one of those things", but I think I will run this past the folks on the S3D forum before I give up. If there is a cure, it's more of a slicer issue than a printer problem.

[ednisley]

layer change is combined with going back & forth

Preview the infill pattern through that section: I suspect a difference directly inside those "defects" which changes the structure just enough to affect the perimeter threads. If so, perhaps picking a different infill pattern might shake the dice enough to produce a less-annoying outcome.

You can see surface changes you can't measure.

[airscapes]

Have you tried to print 2 or more at once? That almost looks like too much heat as the nozzle is always close .. found this to be an issue when printing some really small miniatures .. Can you post the SLT? I have the .5 nozzle on the M2 atm and won't change it for a few days (have to print some Ninjaflex this weekend.. but after that I can put the .25 on and see what it looks like.