Is there any reason on say surface angle > 45 we couldn't halve layer height for external perimeters only?

Subject asks it all

To me it seems like that would weaken the bond between the outer and inner perimeters...

You can speed up your print by using thicker layers for infill than perimeters. For example, if you are using a 0.5 mm nozzle, you can print in 200 um layers and "combine infill" every 2 layers. That will generate infill layers that are 400 um thick and will only print at every second 200 um layer.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Think of it kind of like how stepper motors use microstepping to improve resolution. I want microstepped angles ONLY on surface layers. And for bonus points, variable microstepping -- the shallower the angle to the plate the more micro-steps in external layers.

Slicer already weakens the print by using a thinner external layer to try and get better dimensional accuracy anyway -- so even if halving external perimeter height under those conditions weakened the part anyway for quality, it's already the same tradeoff slic3r is set up to make by default in any case (worse overhang performance with it's default too)...

There are many models I slice at say 0.2mm but say for example the tops of spherical parts would benefit from 0.1mm external perimeter layers for the surface finish (having internal perimeters of 0.2mm would be the bit giving it the strength). With a 0.4mm nozzle I don't want to go much above 0.25 layers or the inter-layer bond strength goes down 0less 'squish'. I don't need the inner perimeters 0.1mm high (though I guess that could complicate things if the externals aren't printed first -- i'm assuming externals first so the internals can 'squish to fill the space' left by the now-overlapping external perimeters). This slicing mode would also seem to benefit from a wider extrusion width than the 1.05 * nozzle default slicer has.

The point of my suggestion is less about speed than quality - though speed is a factor. Printing everything in 0.1 results in lower strength parts (plus it's very slow). And going to 0.05 layers has it's own problems if they were used for everything. Printing "some layers" in 0.1 works fine if you have a nice uniform part where you can slice only the top or bottom of a sphere in 0.1. But on more complicated parts it would seem there's almost no disadvantage to having a slicing rule that performs.

Even taking your example of slicing everything at 0.1 and having 0.2mm infill would seem to gain an advantage in being able to take those very few areas of a model that match the angle-to-the-plane-of-the-nozzle-movement rule and improving the surface finish by adding double the layers. Given that the ideal layer height for strength and other reasons is largely a function of the nozzle diameter and the performance of the heater, I'm unconvinced my suggestion doesn't expand the envelope of speed/quality (even if it's not an enormous expansion, the logic on first blush seems simple enough to codify)

Can anyone point me to the section of the code that makes the slicing decisions? I've had a browse through and it seemed almost like slicing decisions were being made in perl when the docs say it's in c++. I didn't look too hard though. Clues appreciated.

...for visual clarity, this is what I mean:-

[i.imgur.com]

The extra time to print should be incremental if the speed limit is based on the melt-rate of the nozzle anyway as the travel speed will increase to maintain the nozzle flow rate. But the external vertical resolution should increase in multiples. Big benefit for incremental cost seems worth chasing, no?

Also, I've improved my graphic with an alternative depending if it's an overhang or not.

[i.imgur.com]

If I understand your question, this is already something that can be done with some slicer softwares. So the answer is "yes we can".

For instance in skeinforge : [web.archive.org]

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

Is skeinforge the best 'open' slicer then if slic3r can't do "surface skinning"? (as skeinforge calls it... no new ideas under the sun, eh?)

(has it really not been updated since 2012 -- is skeinforge dead or am I just looking in the wrong places?)

Edited 1 time(s). Last edit at 07/19/2016 10:49AM by absd.