There’s been talk recently of printing non-planar layers for (among other goals) getting smoother top surfaces. I had another idea for this. Rather than trying to print lines that don’t follow the X/Y plane, you could lift the nozzle AND increased flow so that you could print on top of a flat layer and maintain the same extrusion width, but the layer thickness would increase in certain areas, up to twice the normal layer height, to create a more gradual transition leading up to the edge of the next layer’s perimeter.
This would mean that movement outside of the X/Y plane would be limited to less than a single layer lift, which would remove most of the worry about hot end and gantry geometry that non-planar layers introduce.
This would mean breaking up printing move segments to include Z axis movement, as spiral vase moves do. Ideally, the speed of these segments should be adjusted to try to keep the extrusion rate constant as the volume per millimeter of horizontal movement increases to (almost) double, so that extrusion speed delay isn’t a factor.
Tried that already years ago. the issue here is that an increased flow will not only result in (somewhat) higher lines, but also wider. Since higher flow also requires a bit higher temperature, they will not only need a longer cooling time before the next layer can be applied, but also extend more in wideth than in height.
Nice idea, but at least I couldn’t get it to work back then.
Well, you have to start with a layer height less than half of your maximum, then lift the nozzle and adjust flow by the same ratio to get the same extrusion width. Just pushing more plastic won’t work. If you’re already pressing the extrusion out to be a little wider than your nozzle, this should work. Also, as I said, rather than pushing plastic faster, you want to slow the X/Y speed to keep the extrusion rate constant.
Sounds like you had close to the same idea, but you were missing at least the feedrate adjustment part. Also, not sure what you mean about cooling time before applying the next layer, because this is only for top surfaces, so there will never be another layer applied over it…
You shouldn’t assume - especiall no negative asumptions.
The cooling is not important when you put the next row. you will notive that the additional widening will be rather dificult.
One solution is to put up the ‘regular’ strips first, and then fill up, starting from the highest fillers first. Still, the result was not as expected.
I finally setteled again for standard methology. Happy to see if you find a way to make it happen.
@Nathan_Walkner No, they’re printing threads of the same cross-sectional area at different heights, with more complicated space-filling algorithms. I’m talking about using existing toolpaths but increasing the cross-sectional area to fill the extra space.
I figure a single outer perimeter prints first (unless one chooses the option not to), then the inner perimeters from the inside out, then work from the stepped perimeters from the outside in? This seems similar to what @Hans_Franke suggested in his last comment, but I think it reads like I’m reversing the order in which you would print the stepped perimeters.
@Tony_Olivo There needs to at least be an option to use (more or less) that order, for hot ends with an overly-large flat on the end, but I honestly don’t think it’s necessary for most.
Most people seem to be using an extrusion width that is too small. Your extrusion width should be larger than your nozzle bore, ideally somewhere between your nozzle bore diameter + layer height and nozzle flat diameter + layer height. If the conical section of the nozzle presses against the taller extrusion a bit as you print a slightly shorter one adjacent to it, that should only enhance the smoothing effect.
Also, keep in mind that for this to work, your layer height has to be less than half of your nozzle bore. Unless you’re using a Volcano, this probably means a layer height no more than .15mm, so those taller lines aren’t going to stick up very far.
@Whosa_whatsis , I think most people let their slicer choose their extrusion widths these days, I know I do :]. So if there’s any trouble with extrusion width too small, I’ll just blame @Daid_Braam .
I also thought most hot ends had moved to a wider nozzle flat rather than narrower. At least my J-head from 2012 has a small narrow nozzle flat that came down as a cylinder from the main taper, whereas my next one in 2013 had no such feature, just a flattened cone.
Unfortunately, Cura seems to put the extrusion width at nozzle diameter. That’s a rather inscrutable decision. I’d really love to see things that can dynamically vary extrusion width for specific tasks, but it gets to be vey complicated too.
@Jeff_DeMaagd It’s a bit more complicated than that. Cura selects an extrusion width based on the nozzle diameter AND the wall thickness. It uses the nozzle diameter to determine the lower limit for extrusion width, but it sets the extrusion width to the nearest multiple of the wall thickness (except in vase mode, where it just uses the wall thickness as extrusion diameter, unless that’s changed). For instance, I set my wall thickness to 1.2mm in Cura to get a .6mm extrusion width for a .5mm nozzle when I’m using .1mm layers.
It may be far simpler to print an object with thick layers but reduce the thickness of topmost layers. Say 0.25 for the whole thing except for 0.1 for the top layers
@Whosa_whatsis I’ve finally had a chance to look at Cura’s line width behavior again, a few days ago. I can try to set the line width to be wider than nozzle diameter, but it colors the box yellow and gives me a warning that tries to discourage that.
The test results turned out to be exactly what I expected, but they seem to have strengthened both my opinions and @Ryan_Carlyle 's about what value should be used.
