[ADDED] Slicer behavior for thin walled sections

[zico]

+1

[dkightley]

Yet another thread calling the software faulty:

viewtopic.php?f=9&t=3611&p=15894#p15894

[bray.christopher]

+

I print original CAD drawings and scaled down, the wall thickness gives lot's of problems. There should be an option to print these thin walls with a minimum of x shells or the same size as the extruder.

Exactly.

Sometimes I am printing a scaled down version of a large assembly for visualization purposes only. Accuracy can take a back seat on these prints. It would take me hours to remodel these assemblies to make them "printable".

Sometimes I am printing a functional part that to test fit, form, and function. These parts need to be accurate first and foremost.

I like the idea of a toggle switch.

[kusmi]

Yes, a configuration setting would be nice. Additionally highlighting problem areas in 3D view or in slicer-view could also come handy - but I'm sure this is not very easy to do

I often run into this problem when I start placing text on my 3D designs. I don't have control on the exact font-metrics, so often some character will print fine, while others get unreadable, because they are missing some walls.
So I often end up using a really large/bold or ugly font, just because one or two characters have problems... I would rather have only a few characters artificially blown up, instead of having 100% ugly text.

[pixel3design]

Hello.
As far as I know, the path solver approach is quite conservative.
That means, if the extruded filament path goes out of the model bounds, it won't be calculated/extruded.

But fortunately there's a workaround.

In the "Extruder" tab, we've got 3 important values:
1. Nozzle Diameter
2. Extrusion Multiplier
3. Extrusion Width (Auto - Manual)

We'll focus on the last one, the "Extrusion Width".
Despite his name, it's just a value that tells to the solver how to behave.
While the first two values could change the extrusion in real world, the third set the virtual space between two paths.

So, putting an Extrusion Width at 0,10 with the Nozzle Diameter at 0,40 and the Extrusion Multiplier at 1,00, will not change the extrusion size in real world, neither change the real volume of the extruded material.

What changes is the solution of the path. In this case u can put 3 shells in a 0,4mm wall using a nozzle of 0,40.

BUT YOU SHOULD BE AWARE THAT THIS WILL LEAD TO THE OVEREXTRUSION, POSSIBLE NOZZLE CLOGS, BAD PRINTS.

I found it not a bug, but instead a solution for small holes and wrong external dimensions due to overextrusion (read Slic3r).

[lciscon]

I will throw my vote in to add an option to allow the path to exceed the shape slightly. The approach of reducing the extrusion width does not produce the same result and is much more likely to cause jams. I have tried this approach and found it to be unreliable.

Note that this is not just a "minor" issue. I've encountered this problem on several print jobs and will be forced to use a different slicer software tool for certain projects in the future unless this is resolved.

One comment on those suggesting you go back and modify the original file: Many of us are now trying to make money on 3D printing and are running print jobs for other people. As such we may not have access to the original CAD model file. Moreover "growing" an existing STL file is tricky and tends to produce a lower quality output.

[ashleywebster]

I agree with the general consensus. More choice = better. I will just stress that the user should be warned in some way that this will impact dimensional accuracy. It is imperative that we can trust S3D to create accurate parts when we need them.

Pixel3Design is correct in that you can fudge it with those three settings. But extrusion width most definitely changes things in the real world - all of them do. Layer height and extrusion width should be on the same tab because together they determine the cross sectional area of the thread of plastic. A higher or lower extrusion width will change the volume of plastic being extruded. I'm printing a spiral vase right now and have the extrusion width set to 0.85mm because I want the single perimeter to be fairly thick. And sometimes I set it below the real nozzle width to force small features to print. The risk of going below the nozzle width is that the plastic won't get squished down well and your layer bonding will be weaker.

The fixes that have been recommended are definitely superior to the current workaround.

[CleverCoder]

I'm going out on a limb here, and guessing that this is the reason I'm unable to print an 0.5 mm thin walled test cube:
http://www.thingiverse.com/thing:1637

Using 0.4mm extruder, 0.5mm extrusion width, 1.0 multiplier, in the preview window the cube simply doesn't appear. This is with my FlashForge Creator Pro. Other stuff seems to print fine when the width is greater than some magical size.

I was going to start tinkering with settings, but felt it was a losing battle.

- Sean

[CleverCoder]

Okay... I wanted to update my post. It looks like the 05.mm test box I linked to loads up in Netfabb, and I'm measuring 0.24 mm wall thickness!
This actually makes much more sense as to why an 0.5mm extrusion width would refuse to attempt printing.

[jdub1981]

Not sure if the issue i WAS having is the same as the op's but by simply adding .1 to the horizontal size compensation and then previewing again I was able to get the very thin pieces of my part to preview correctly and then print correctly.