Page 2 of 12
Re: Slicer behavior for thin walled sections
Posted: Tue Sep 15, 2015 11:59 am
by wfcook
Fred B wrote:Cross sections below the extruder width don't currently print. This shouldn't be a surprise because we want accurate prints.
The other slicers that "force" thin walls to print will sacrifice accuracy. Why would we want to print inaccurate parts? It's more accurate to design features to print properly in the first place.
If they did add a feature it could be called "inaccurate thin wall mode"
The current behavior, where the slicer leaves a large gap in the part, is not a good solution either. It is not always possible to simply redesign the part to solve this problem, as the part requirements may not be compatible with a single, constant extrusion width. Simplify3D offers no variation in extrusion width for different elements of the print and no dynamic-width gap fill. To make this work, it needs both.
In contrast, Slic3r (which has its own issues, to be sure), nicely lets you vary the extrusion width for outer perimeters, inner perimeters, infill, etc.
Re: Slicer behavior for thin walled sections
Posted: Wed Sep 16, 2015 5:45 am
by dkightley
Slicing this test part illustrates the whole set of issues:
http://www.thingiverse.com/thing:30340
Re: Slicer behavior for thin walled sections
Posted: Thu Sep 17, 2015 4:17 pm
by tkatsouras
My choice here is to print thin walls that are smaller than the diameter of the nozzle with one line and get a warning from the S3D software.
I also like the option to print thin walls with lines instead of 1 or 2 lines and have infill, that makes the print cleaner, I understand there will be gaps sometimes when when the thin wall width is not multiple of the nozzle diameter, that is OK with me since I use very small nozzles and the gap if any will be very small.
Ted
Re: Slicer behavior for thin walled sections
Posted: Mon Sep 21, 2015 12:03 pm
by Rebekah_harper
I think that as designers and operators who have a decent grasp of what we want that we should be able to choose to either ignore wall limitations or to accept them.
surely, as suggested, a tick box allowing the operator to choose to either have the accuracy and the design flaws highlighted or to sacrifice the accuracy for appearance.
I would like to have the choice as there are times where I just want it to print simply because the feature I have issue with is just decoration and doesn't require accuracy.
I vote for freedom of choice.
Becky
Re: Slicer behavior for thin walled sections
Posted: Mon Sep 21, 2015 7:50 pm
by tonyno
I agree with that.
Re: Slicer behavior for thin walled sections
Posted: Tue Sep 22, 2015 12:43 pm
by gatorNic
Rebekah_harper wrote:I think that as designers and operators who have a decent grasp of what we want that we should be able to choose to either ignore wall limitations or to accept them.
surely, as suggested, a tick box allowing the operator to choose to either have the accuracy and the design flaws highlighted or to sacrifice the accuracy for appearance.
I would like to have the choice as there are times where I just want it to print simply because the feature I have issue with is just decoration and doesn't require accuracy.
I vote for freedom of choice.
Becky
+1
There are a lot of people out there who aren't actually modeling parts, they are downloading off of thing-verse and just want it to print without worrying about exact dimensional accuracy. Even myself as a professional modeler, could pretty easily fix other people's models. However as I just mentioned there are times when I just want to print something and not spend lot of extra time fixing when dimensional accuracy isn't a concern.
Re: Slicer behavior for thin walled sections
Posted: Wed Sep 23, 2015 8:27 am
by dkightley
From my original post...which actually raised two issues:
The second issue relates mainly to the quality of top & bottom surfaces where the width of the wall section is not a multiple of 2 times the extruded filament width...resulting in gaps and holes on the surface. The "Allow gap fill where necessary" option should get rid of the gaps....but it doesn't seem to do so reliably....and it certainly doesn't show up as having much of an effect on the print preview. A more effective gap filling control is required...preferably one that does not need adjustment of several other options as well.
Here's an idea that may help to sort out top surfaces...
Currently, the slicer lays down perimeter layers (however many there has been specified) and then lays down solid fill (if any) within the perimeters. Why can't the slicer be made to not lay down the second and subsequent perimeter whenever it is going to place a solid layer (but not infill). This then may give the slicer the chance to make a better job of filling small gaps...particularly on thin sections!!
Re: Slicer behavior for thin walled sections
Posted: Wed Sep 23, 2015 11:12 am
by CompoundCarl
dkightley wrote:Currently, the slicer lays down perimeter layers (however many there has been specified) and then lays down solid fill (if any) within the perimeters. Why can't the slicer be made to not lay down the second and subsequent perimeter whenever it is going to place a solid layer (but not infill). This then may give the slicer the chance to make a better job of filling small gaps...particularly on thin sections!!
You literally described exactly what the existing gap fill algorithms in the software already do. The software does NOT always lay down EVERY single perimeter you specify. If you enable gap fill, it already intelligently switches to a variable width fill to cover that gap. Just tested it myself. It definitely already does this.
Re: Slicer behavior for thin walled sections
Posted: Wed Sep 23, 2015 11:52 am
by dkightley
Allow me to explain what I mean
I've set up two cubes, one set with 5 perimeters and the other with 1. Here's the preview almost at the top:
The cube on the left has the 5 perimeters with the honeycomb infill. The cube on the right ditto, but only one perimeter.
One layer later:
The solid layer that forms the top has printed INSIDE the perimeter. What I am suggesting is that at this point, both cubes should look like the right hand cube. IE the solid layer is printed instead of perimeters 2 to 5. Thus the top surface will not bear the scars of the inside perimeters.
Re: Slicer behavior for thin walled sections
Posted: Thu Sep 24, 2015 12:13 pm
by wfcook
CompoundCarl wrote:
You literally described exactly what the existing gap fill algorithms in the software already do. The software does NOT always lay down EVERY single perimeter you specify. If you enable gap fill, it already intelligently switches to a variable width fill to cover that gap. Just tested it myself. It definitely already does this.
I am not seeing this behavior at all, please share whatever settings make it happen. In my tests, S3D either leaves out the gap fill (if the space is less than the extrusion width) or fills using the infill pattern (with the attendant vibration, noise, and other issues). I do not see any way to make it dynamically change the extrusion width.