A wierdo - any ideas?

[yellowshark 153]

I am printing this part; this pic shows the orientation on the print bed. As you can see quite a severe overhang, I am not using supports.

My initial settings were; .300 layers, 40 mm/s, 100% fan, 210c extruder, 0.4mm nozzle, 20% infill and 1.2mm wall thickness. After my first run I had the obvious idea to swap the nozzle to 0.8mm to reduce the print time and cost. .

The print came out quite well, better than I thought it might do – this pic is looking at the overhang from underneath –

 

I started another print to check it with my grey filament and I noticed that in Cura I still had the wall thickness set to 1.2mm from using the 0.4mm nozzle of the first print, so I changed it to 1.6mm (everything else was the same)and started the print. I got this on the underneath of the overhang.

I cancelled the print, changed the wall setting back to 1.2mm, and printed again; just a couple of minutes in-between prints. I got this

 

Now I am not sure if this is a learning experience but I have no idea what is happening here. I have always thought it is best to make the wall width a multiple of the nozzle width, but this would suggest that is not true! Any ideas what was happening?

 

 

 

[gr5 2,069]

Woah - I think you just blew my mind! :eek:

Let me summarize as you tried to be very exacting but want to be sure. There are 4 photos in your post:

photo 1) .4mm nozzle (real) .4mm nozzle (cura) 1.2mm wall (green)

photo 2) .8mm nozzle (real) .8mm nozzle (cura) 1.2mm wall (green) (Is this 2 0.6 passes? Or one 1.2mm pass? I'd have to check the code in cura)

photo 3) .8mm nozzle (real) .8mm nozzle (cura) 1.6mm wall (gray) UGLY

photo 4) .8mm nozzle (real) .8mm nozzle (cura) 1.2mm wall (same settings as photo2 but with gray filament)

Is that all correct? Could you double check?

If that's all true then the lesson perhaps is: if you print with really thick layers (.3) and fat traces (.8) it takes longer to cool and overhangs look crappy and this is more important than getting the wall width a multiple of nozzle width.

 

[yellowshark 153]

Thanks for that George, nice to know someone else’s brain is frazzled :-P . Ok pretty much spot on, a couple of clarifications.

Photo 1) 95% sure that was of version printed at .8mm nozzle real and in Cura

Photo 2) I did two more prints, just a few layers with a 1.2mm wall and a 1.6 mm wall, to check.

The 1.2mm setting produces one 1.2mm pass – it measures 1.35mm but given the few layers printed and I ripped it off the bed so it is bent, that measurement may not be too accurate.

Yes the idea of single pass versus double pass did go through my mind although not as analytically as went through yours. Certainly the fans were on 100% once the slope started. The first pass will be partially seated on the previous layer, I assume, which will be cooled because of the 20% infill work before starting the next layer. Again I assume, that the 2^nd pass adheres to the side of the first pass, or not!

The very first run I did was 0.4mm with 1.2mm wall, using supports added by Meshmixer. The surface on that came out well also, but breaking off the supports just left me with too much work in filing down the break points to get a smooth finish ready for painting.

I am thinking now that maybe the way to get really well finished overhangs is to go with a single pass. I have some pieces I did a year ago where I struggled to get a decent finish on the overhang (I have always used at least double pass walls) and I think I will revisit them with this new theory :cool:

 

[foehnsturm 969]

Single pass is not better as a matter of principle. But if the nozzle comes back to the same location before the extruded filament cooled down sufficiently, you are in trouble. 0.8 x 0.3 mm2 is a huge cross section, it will take some time to cool down.

Single pass with infill (and a minimum layer time setting) will always avoid an untimely return to the same location. With a double pass there is absolutely no way to control that besides print speed: time to return = perimeter length / print speed.

 

[yellowshark 153]

Hi Foehnstrum, thanks for that. I wonder if you have typed a word wrong somewhere. If I read your 2nd paragraph first, then surely your first paragraph should say - " Single pass is better as a matter of principle"?

 

[foehnsturm 969]

Thanks, it should read "Single pass is not naturally better".

What I tried to say is that if the part is big enough a double pass is at least on par with a single pass. Perhaps it even has some advantages.

 

[yellowshark 153]

OK I understand and I agree. I cannot prove it, as I have never done any comparative testing, but with a 0.4mm nozzle I have always felt that multiple passes, versus one pass, reduce the risk of the exterior surface being affected by the infill process, particularly on circular columns; where I have improved the surface by adding another pass.