Bridge speed and flow ratio is used on incorrect solid infill layers.

[TheBoojah]

Version

1.2.9 and latest dev (oct 23)

Operating system type + version

Win 10 latest patch

Behavior

Bridge speed and flow ratio is used on incorrect solid infill layers.
Bridge should be use on the lowest top-surface solid layer, i.e. the first layer on top of the infill.
What I am getting is all solid layers except the very top and bottom ones use bridge speed and flow. So for 3 bottom 3 top, the first is ok, then 2 bridge layers - infill - 2 bridge and the final top layer.

To illustrate I take all the default settings, except changing bridge flow ratio to 0.1 to make it more visible:

The solid infill lines are very thin, indicating that bridge settings were used.
This is just a 20x20x10 cube, and I believe there is no other settings changed in Slic3r - everything is default.

If this is not easily reproducible please let me know what additional info I can provide.
Thanks

Update, model and config:
#4172.zip
The model is simply a small box and this issue is present on all the models I have tried.

[lordofhyphens]

Please zip that model and an export config (Ctrl-E) and attach it to the issue (drag and drop)

[TheBoojah]

I updated the issue with the requested files.

[lordofhyphens]

As far as I am aware, the first "solid" infill over sparse infill is bridge settings by design (as that's really what it is, a bridge).

I can see a case for the bottom solid layers (which are handled somewhat differently) not being treated as solid infill.

Is this causing a printing issue for you (this is curiosity on my part).

[lordofhyphens]

It appears that all internal solid infill is being flagged as bridges, beyond the first layer above sparse infill.

It's definitely wrong for bottom layers that aren't above a void.

[lordofhyphens]

Other notes:
At 100% infill bridge infill is not used at all, so there's a special case going on here.

If you wanted to have a workaround for a simple case you can set the main mesh to 100% infill and make a modifier mesh that is scaled down slightly and set to the real infill density. Slic3r does not seem to pay attention to infill differences for modifier meshes for the purposes of internal bridge infill.

[lordofhyphens]

I am pretty sure the error is in the surface type detection. Top and bottom surfaces are easy to find, so those are classified correctly.

The associated code should be in Layer::detect_surfaces_type().

It's actually applied around line 205 of LayerRegionFill.cpp where the flow is calculated.

[TheBoojah]

Thanks for looking into this bug!

The way I detected this is playing with the bridge settings, at 0.8 bridge flow ratio the thinner lines starts to get really noticeable in the preview. The real killer is however when it does 10mm/s on the second solid layer of a larger part. So I can se why nobody has reported this to far, as when you have default or "light" bridge settings you never notice a difference. On challenging parts with some more extreme bridge settings can really mess things up, a high bridge fan setting on a large second solid layer is not so great for bed adhesion for example.

Thanks for looking for a workaround. It won't work for my current model so another way is just to use close to default bridge settings and hope the bridges come out ok anyway.

I tested Slic3r 1.1.2 which worked correctly, only on the first solid layer over sparse infill is bridge settings used.

[lordofhyphens]

Yeah, this is a regression from when the logic was ported.

…

On Oct 28, 2017 4:04 AM, "TheBoojah" ***@***.***> wrote: Thanks for looking into this bug! The way I detected this is playing with the bridge settings, at 0.8 bridge flow ratio the thinner lines starts to get really noticeable in the preview. The real killer is however when it does 10mm/s on the second solid layer of a larger part. So I can se why nobody has reported this to far, as when you have default or "light" bridge settings you never notice a difference. On challenging parts with some more extreme bridge settings can really mess things up, a high bridge fan setting on a large second solid layer is not so great for bed adhesion for example. Thanks for looking for a workaround. It won't work for my current model so another way is just to use close to default bridge settings and hope the bridges come out ok anyway. I tested Slic3r 1.1.2 which worked correctly, only on the first solid layer over sparse infill is bridge settings used. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#4172 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAB8CigOdsG75yzpW_ftRVn0wJqYTO_5ks5swu4GgaJpZM4QIIbF> .

[wirlybird]

Hey guys, came across your posts searching for a solution to a similar issue. This pic is the first layer over a support. It is being treated as a bridge and making the extrusions too thin so they mainly do not attach to anything and when you remove the supports you have a stringy mess.
Increasing the flow percent just pushes the extrusions further apart so the gaps stay the same and it doesn't solve anything.
Now, I can understand why it may think it needs bridging here but it just makes a mess.

If I run the same part in S3D it builds a solid layer over the support and it is all good.

Would love to find a solution! Even just being able to completely disable bridging would be good.