"Tree-branching" when printing multiple small parts

[ChrisERAU]

So I just checked the actual tessellation of the prop blade. It doesn't look too busy (at least to me). But it's something I'll keep an eye on for future models. I do rendering on the side, so high tessellation is just a habit

If you're going to use only 15% infill, you want to use at least 4 or 5 top layers. (Those holes sometimes form over skimpy infill, you need enough top layers to make sure it gets filled in.) Am I printing the infill too fast? It's on the default 80mm/s, and it's fine when it lays down the layers on the build plate, but it gets iffy when it prints on top of any infill.

Bottom layer looks fine - might be a bit over-extruded, but that's better than not enough.

Well that's the thing... I did use 5 layers of top infill. I confirmed in S3D's preview that all 5 layers were there. And yet I still get the holes.

I have something printing now, but I'll get to the hollow box test model afterwards. I tweaked some settings for the current thing printing: 8mm/s outlines, 25% fast hex infill, and .2mm vertical ret. lift for .1mm layer height. It's doing great so far, very very little ghosting around the corners ( ). We'll see how the top layer comes out.

By the way, here's the rest of the plane I'm printing. Only the wings left to print. It was my group's preliminary design senior project last year. I enjoy CAD so it was my baby for a semester.

[ednisley]

It doesn't look too busy (at least to me)

That's because your eye is calibrated for visual effects, not 3D printing! [grin]

If the prop blade is 25 mm long, then the chord is maybe 2 mm near the middle, so the total perimeter might be 3 mm or so. More than three dozen triangles along those nearly horizontal lines doesn't improve the results; I lost count long before that.

At the base, you have grossly exceedingly far too many excessive segments around that 2 mm circle. Two dozen would be plenty; you'll never see the difference beyond 16.

Really and truly, lower model resolution will avoid many, many subtle printing problems, without compromising the visual quality of the final plastic shape.

it gets iffy when it prints on top of any infill

Slow down the top layers to maybe 1/3 of the internal infill: you want quality more than speed.

Rather than mess with the individual speeds, get it working properly on larger models (like the nacelles), then adjust the minimum layer time for smaller objects. That scales all of the speeds proportionally, gets you out of all the twiddling, and lets you concentrate on the Big Picture.

With that in mind, if the first layer of larger objects is jam-packed full and the top layer is skimpy, that tells you to pay attention when you measure those five thinwall hollow boxes: the initial platform gap isn't quite right and you may not have the Extrusion Multiplier set.

If a Makerbot can do it, I think the M2 should, right?

That's correct, but unless you devote some time to optimizing the process variables for your particular models, either printer will produce complete junk. The settings won't be the same from one printer to the next, particularly printers with such different mechanics, so you can't assume what works for one printer will work for another.

[ChrisERAU]

Squares printed pretty well, or at least as expected. The middle square had the best adhesion and the best first layer... and consequently the lowest height. The top left square was a close second, and the other outer ones were about the same: weaker adhesion and not quite squished first layers. I measured the thicknesses (in blue) and the sides (in yellow). Pics below. I also added my print settings. What can I do to minimize the apparent central bowing effect?

Edit: I forgot to add the wall thicknesses in the pic. They ranged from 0.7 to 0.85 mm. 0.8 mm was the most common thickness. I confirmed in S3D that the walls were one filament wide, yet they printed quite a bit over the expected 0.4 mm....

[ednisley]

at least as expected[/attachment]

OK, let's recalibrate your expectations, too... [grin]

If you've set the slicer to produce 0.40 mm walls and you're measuring 0.8 mm walls, then your settings are off by a factor of two: the walls should be exactly 0.40 mm. You're not going to get anywhere until the printer's extruded plastic matches the slicer's instructions.

Go through the Extrusion Multiplier calibration routine:
viewtopic.php?f=3&t=1964

Pay attention to measuring the actual filament diameter, because a factor of two error seems unusually large.

When you've iterated on that and gotten good results, slice & print another set of five thinwall cubes (which will now have exactly the right wall thickness), and measure their top-to-bottom height. That will lead you into the platform alignment fine-tuning routine:

Coarse:
viewtopic.php?f=3&t=2783

Fine:
viewtopic.php?f=3&t=3047

When you're getting exactly the right heights, then things should start looking better all over... honest!

For those of you reading this in our future, you should start with Jules' guide, because it lists nearly everything you need to know in one place, with helpful commentary:
viewtopic.php?f=8&t=2778

[Pekish79]

i had that happen to but it was because i forget to fix the "2 layer" perimeter so even if i draw 1 layer he still printed to layer so i think his .8 is correct as they are 2 layers

honestly is almost impossible to make a layer of .8 with a single pass

[ChrisERAU]

i had that happen to but it was because i forget to fix the "2 layer" perimeter so even if i draw 1 layer he still printed to layer so i think his .8 is correct as they are 2 layers

honestly is almost impossible to make a layer of .8 with a single pass

Well, prepare to be amazed. I printed one square again, and made sure to watch it do about 5 layers, and it's definitely a 1 filament wide perimeter. Just checked the previous squares, and they were 1 filament perimeters too. My printer settings were mostly from the M2 profile that was included with S3D, with tweaks only to infill, start/end scripts, and print speed, along of course with layer height. I didn't mess with multipliers, nozzle width, or extrusion width. Pretty baffling. I printed another square at .3mm layer height, and the dimensions remained pretty much the same.

And wait, there's more: the supports are only 0.4mm wide. Huh?

[Pekish79]

not sure how is possible i had it .8mm before but it was always before i forgot to fix the gcode
are you using a full box or make a perimeter only box?

perimeter only box then to double the thickness eve if u tell them to do 1 layer as they count left face right face as 1+1
if you make a full box and tell no infill no top no bottom then you really get just 1 layer

[ChrisERAU]

not sure how is possible i had it .8mm before but it was always before i forgot to fix the gcode
are you using a full box or make a perimeter only box?

perimeter only box then to double the thickness eve if u tell them to do 1 layer as they count left face right face as 1+1
if you make a full box and tell no infill no top no bottom then you really get just 1 layer

Perimeter only box, same as the ones I posted in the pic earlier. I did make a discovery, however, when preparing to do the extrusion calibration Ed mentioned. Someone said to "merge all outlines into a single model". When I did that, the print preview looked exactly the same as before, but the estimated filament weight and length usage dropped by about half. This implies that the walls will now only be 0.4 mm thick.

I then reprinted the square at 0.3 mm layer height, and sure enough.... the walls were now 0.44 mm thick. The height was 5.03 mm.

I reprinted yet again, changing only the merging parameter, and I get the 0.8 mm walls and 5.19 mm height like before.

Conclusion: Merging in S3D is very wonky. Now to run the extrusion calibration with merging parameters enabled!

EDIT: I ended up with an extrusion multiplier of 0.82. Is that normal? The square looked the same as the 0.44 mm thick one (0.04 mm is imperceptible just by glancing), so I guess yay, I get to use slightly less filament now?

[Pekish79]

yeah i had something like that happen as well but i was sure that everytime i had .78 or .82 was 2 layer no matter what S3D was saying

so i tried different stuff since i convinced it to print .4 that why i was sure u had 2 layer because the program doesnt show it right but is pratically impossible to have .8 with a .35 nozzle in single pass

[ChrisERAU]

So, now going all the way back to the original problem , I'll see if I can print the propellers correctly. I'll start with one, as recommended earlier, and go from there.

Also, it looks like someone else encountered this "glitch", if I can call it that. https://forum.simplify3d.com/viewtopic.php?f=9&t=2217