S3D Strange Slicing

Have questions or comments about Simplify3D, Slic3r, Cura, Reptier, etc? Or wondering about which CAD software to use...discuss it here...
Toby
Posts: 330
Joined: Wed Apr 09, 2014 9:44 pm

Re: S3D Strange Slicing

Post by Toby » Wed Apr 16, 2014 3:38 pm

I'm not convinced that S3D doesn't have any bugs in the slicer. It does strange thing sometimes, and the fixes often don't make a lot of sense- like translate the model on the bed and slice again.

As was pointed out before (by Dale I think), floating point round-off could be the culprit. The funny thing about STL files is that they don't have connectivity information. The only way the slicer knows two triangles share an edge or a vertex is because of their floating point values . However, this can get skewed, for example when slicing through two triangles that share an edge. The point where the slicing plane hits one edge may be different from where it hits the other edge in their floating point representations depending on how the slicer orients the triangles while slicing. This can lead to the slicer thinking there's a gap, or there's no closed polygon, or something else that causes it to fill in some areas or leave other areas out. To fix it the slicer has to be intelligent about "near enough for printing", which I think is why the "heal" method often works.

BTW, don't forget that the first layer height can have an impact on adhesion. I'd go down to 90% rather than up to 110%.

jsc
Posts: 1864
Joined: Thu Apr 10, 2014 4:00 am

Re: S3D Strange Slicing

Post by jsc » Wed Apr 16, 2014 4:50 pm

Just try different first layer heights and make sure you check out the gcode preview from the top view to make sure the holes haven't been closed. You don't need to adjust by 10%, 1% is fine. You are just trying to slightly shift the layers to avoid problem geometry, after all. The smallest change should work. Alternatively, you can try changing your layer height by a tiny fraction, so instead of 0.2mm, you could try 0.19999.

Regarding the filament hole, you mean 2mm, not 20mm, right? Your filament is 1.75mm, so that should be plenty of clearance.

Zeeflyboy
Posts: 21
Joined: Thu Apr 17, 2014 3:31 pm

Re: S3D Strange Slicing

Post by Zeeflyboy » Thu Apr 17, 2014 8:37 pm

also regarding filament holes, it's always best to skim down them with an appropriately sized drill bit I'm my experience.

3D printers aren't great at perfectly size and smooth small holes or tubes, so often benefit from a little "post processing".

for what it's worth I've had similar extremely annoying errors in S3D too with models I've produced in solidworks.

User avatar
jimc
Posts: 2888
Joined: Wed Apr 09, 2014 11:30 pm
Location: mullica, nj
Contact:

Re: S3D Strange Slicing

Post by jimc » Fri Apr 18, 2014 3:43 am

well just for the hell of it i downloaded silvers filament drive and sliced it with s3d and it sliced perfectly. no issue at all.

Toby
Posts: 330
Joined: Wed Apr 09, 2014 9:44 pm

Re: S3D Strange Slicing

Post by Toby » Fri Apr 18, 2014 5:19 am

S3D is giving me fits today. I have 26 parts I need to print to construct a model.

One part is just a walled box with the top removed. I have to print 4 of them. I get that random total infill depending on I-don't-know-what. Change the layer height, this one is fine, that one is filled. Or sometimes they all get filled for just a portion in the middle. That fooled me into printing because it's not obvious in the gcode preview. But half way through suddenly I'm getting layer after layer of solid fill.

Translate on the bed- same thing. Sometimes it's good, sometimes not. The only thing I found that was reliable was to rotate each one a little bit around the z-axis.

This is mondo frustrating. At the moment I'm cursing S3D as not being fit for production printing. Fine if you just want to print a single item and don't mind fiddling with parameters to get it right. Not good if you have a bunch of things to print and you end up having to fiddle with each one and keep track of all the little changes across each model. And then if you have to change some models or settings- do it all again. Endless mouse clicks because of poor file management. Not good not good not good.

The M2 meanwhile is flawless. Print after perfect print.

Zeeflyboy
Posts: 21
Joined: Thu Apr 17, 2014 3:31 pm

Re: S3D Strange Slicing

Post by Zeeflyboy » Fri Apr 18, 2014 6:26 am

I'm using slic3r more and more... Not least because I find the way they do profiles much more intuitive than s3d.

Slic3r also seems much more intelligent when it comes to bridging and anchors much better... Honeycomb is also the best infill pattern IMO and overall I'm getting better end results from slic3r. It does occasionally do some strange stuff too but it's much less common in my experience and usually related to very high resolution STLs.

I do still use S3D a lot to preview the Gcode as it has an excellent preview function.

jsc
Posts: 1864
Joined: Thu Apr 10, 2014 4:00 am

Re: S3D Strange Slicing

Post by jsc » Fri Apr 18, 2014 7:46 am

Send the factory file to Clayton and complain. It shouldn't be messing up on a box! That sounds more like an actual bug that "difficult geometry" causing problems.

roberts
Posts: 8
Joined: Fri Apr 11, 2014 2:36 am
Location: Miami, FL

Re: S3D Strange Slicing

Post by roberts » Fri Apr 18, 2014 11:53 am

jsc wrote:Just try different first layer heights and make sure you check out the gcode preview from the top view to make sure the holes haven't been closed. You don't need to adjust by 10%, 1% is fine. You are just trying to slightly shift the layers to avoid problem geometry, after all. The smallest change should work. Alternatively, you can try changing your layer height by a tiny fraction, so instead of 0.2mm, you could try 0.19999.

Regarding the filament hole, you mean 2mm, not 20mm, right? Your filament is 1.75mm, so that should be plenty of clearance.

I did finally get the guide to print near perfectly with a little fiddling with first layer height and orientation to get S3D to stop doing things it really shouldnt be. The filament hole did seem a tad snug but was able to clean it out well enough to allow for smooth operation. I've been printing with the new guide for a couple days now and its working well. I really wanted that receptacle at the top to hold the tube as I've had some filament break overnight laying against the open hole due to gravity and/or bend from being uncoiled. With the tube supporting the filament itself it stays perpendicular to the hole and any stress is on the much stiffer tube. It working great so far. Also my first experience taking apart that whole assembly which was good learning experience, it went well.

Sorry for the confusion on the 20mm, I was referring to the height of the overall part. I was just checking to see if my part printed was close to the actual model. It came out as 19.88mm vs the 20mm model size which is close enough it seems.

Thank everyone for your help.

Toby
Posts: 330
Joined: Wed Apr 09, 2014 9:44 pm

Re: S3D Strange Slicing

Post by Toby » Fri Apr 18, 2014 8:52 pm

jsc wrote:Send the factory file to Clayton and complain. It shouldn't be messing up on a box! That sounds more like an actual bug that "difficult geometry" causing problems.
I changed the design slightly last night and the problem went away. But I think I will try to reconstruct it and forward it to S3D. It was like a total of 20 polygons that passed all tests in netfabb and s3d, but sliced bizarrely depending on where it was located on the bed.

Post Reply