Ultimaker / Cura

3D printer / slicing GUI built on top of the Uranium framework
GNU Lesser General Public License v3.0
6.12k stars 2.06k forks source link

Slicing error when using higher resolution print, OK for lowest resolution. The submission form does not allow me to attach the .3mf filettach the #17103

Open Yorkshire-Farmer opened 11 months ago

Yorkshire-Farmer commented 11 months ago

Cura Version

5.5.0 - Beta 1

Operating System

windows 11

Printer

Ultimaker S5

Reproduction steps

Load STL file (file loaded and schematic appeared on build plate as usual) Set nozzle material to Tough PLA, with no support Set slicing operation Slicing failed with error message suggesting that I submit a error report

Actual results

Nothing the design image remained on the build plate and slicing option re-appeared, when pressed the same results occurred

Expected results

The sliced design should have become available to be transmitted to the printer.

I tried a variety of options, the only one that worked was to have the ultrafast print (0.3mm) with no support and no adhesion

Add your .zip and screenshots here ⬇️

I tired to attach the .3mf project file but it was rejected as an unsupported file type

GregValiant commented 11 months ago

Thanks for the report. Github rules are that anything other than a text file must be zipped. We'll need that project file ("File | Save Project") to investigate. If the file is deemed too large by Github you can post it on Google Drive (or something like it) and leave a link here.

Yorkshire-Farmer commented 11 months ago

Dear Greg,

herewith the zip version of the project file

On Wed, 25 Oct 2023 at 23:42, GregValiant @.***> wrote:

Thanks for the report. Github rules are that anything other than a text file must be zipped. We'll need that project file ("File | Save Project") to investigate. If the file is deemed too large by Github you can post it on Google Drive (or something like it) and leave a link here.

— Reply to this email directly, view it on GitHub https://github.com/Ultimaker/Cura/issues/17103#issuecomment-1780156716, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDKFWH4UX2N34OU4A7XLWDTYBGIVBAVCNFSM6AAAAAA6PM5WGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBQGE2TMNZRGY . You are receiving this because you authored the thread.Message ID: @.***>

-- Dr. Peter Bailes P & C. M. Bailes Farming High Inhams Farm Stillington Road Sutton on the Forest York North Yorkshire YO61 1EQ Tel 01347-812589 Fax: 01347-811827 Mobile: 07796622921 Email: @.*** GPS 54 °04' 39.0" N 1°06' 11.3" W (54.0775°N,1.1030556 °W) We are what we repeatedly do. Excellence, then, is not an act, but a habit.

GregValiant commented 11 months ago

You have to post the zip file here directly. It doesn't come through when trying to use Email. It's a GitHub thing.

Yorkshire-Farmer commented 11 months ago

UMS5_Gully_Linear_Pattern_V3-Fillet001.zip

Yorkshire-Farmer commented 11 months ago

Hopefully this has got through

Yorkshire-Farmer commented 11 months ago

Dear Greg,

I have sent the file using the git hub link, hopefully it has arrived with you in a usable form.

Best Wishes

Peter

On Thu, 26 Oct 2023 at 17:09, GregValiant @.***> wrote:

You have to post the zip file here directly. It doesn't come through when trying to use Email. It's a GitHub thing.

— Reply to this email directly, view it on GitHub https://github.com/Ultimaker/Cura/issues/17103#issuecomment-1781425521, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDKFWH7M6QK4NXMG2ZDUDS3YBKDM3AVCNFSM6AAAAAA6PM5WGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBRGQZDKNJSGE . You are receiving this because you authored the thread.Message ID: @.***>

-- Dr. Peter Bailes P & C. M. Bailes Farming High Inhams Farm Stillington Road Sutton on the Forest York North Yorkshire YO61 1EQ Tel 01347-812589 Fax: 01347-811827 Mobile: 07796622921 Email: @.*** GPS 54 °04' 39.0" N 1°06' 11.3" W (54.0775°N,1.1030556 °W) We are what we repeatedly do. Excellence, then, is not an act, but a habit.

GregValiant commented 11 months ago

I got it. When Cura shows "red surfaces facing upwards" it indicates surfaces that are "inside out". WHen the slicer encounters them it can get confused. Here is a view of the single rib. This does not look right and is not repairable. image

This is one of the short ribs and is also not repairable. image

Here is that same short rib after repair by MS 3D Builder. image

After moving the model on the Cura build plate it sliced, but I don't think this is what you had in mind. This is the short rib image

The model has non-repairable errors and so we can't tell if a bug is affecting slicing, or if the model is affecting slicing.

I see you have closed this report. Is that what you wanted to do? It's OK since there isn't anything we can do with this because of the bad model.

Yorkshire-Farmer commented 11 months ago

Dear Greg,

Thank you for this analysis. Having now printed this, I have also observed the same thing, but it does leave a few points hanging.

The first is that the latest Cura software does manage to slice this identical model, when the resolution is decreased to a 0.3mm feed (I have not printed this), the second is that if I revert to the previous Cura release (5.3), it does successfully slice the model at 0.1mm feed.

I tried this and also printed it and the result is fine except for these support brackets, which , as you illustrate, do not come out as solid supports. They are rendered in Freecad as solid supports, but clearly something is missing in the translation. They were formed by creating a "loft" between two identical closed 2D shapes attached to the horizontal and vertical faces of the gulley cover. I will do some more research on what could have gone wrong in the configuration of the "loft", but it is worth noting that Cura version 5.3 did allow the slice to complete and also did allow the model to print to completion (albeit with faulty support brackets).

I will let you know if I find a root cause at the freecad side of the project or the "export" of the stl file, for your awareness in dealing with other users.

In the mean time many thanks for your support and the effort that you have devoted to understanding this problem.

Best Wishes

Peter

I

On Sat, 28 Oct 2023 at 12:05, GregValiant @.***> wrote:

I got it. When Cura shows "red surfaces facing upwards" it indicates surfaces that are "inside out". WHen the slicer encounters them it can get confused. Here is a view of the single rib. This does not look right and is not repairable. [image: image] https://user-images.githubusercontent.com/64202104/278809023-befc5ed4-4c59-4ee8-a572-5fc2657e903d.png

This is one of the short ribs and is also not repairable. [image: image] https://user-images.githubusercontent.com/64202104/278809111-0856e71a-c56c-4a43-aab8-c569e0a1fe9d.png

Here is that same short rib after repair by MS 3D Builder. [image: image] https://user-images.githubusercontent.com/64202104/278809197-6c0f5408-8680-4d54-a805-46763c54ce55.png

After moving the model on the Cura build plate it sliced, but I don't think this is what you had in mind. This is the short rib [image: image] https://user-images.githubusercontent.com/64202104/278809364-0db7e396-1a8d-402d-aa4a-c330682db637.png

The model has non-repairable errors and so we can't tell if a bug is affecting slicing, or if the model is affecting slicing.

I see you have closed this report. Is that what you wanted to do? It's OK since there isn't anything we can do with this because of the bad model.

— Reply to this email directly, view it on GitHub https://github.com/Ultimaker/Cura/issues/17103#issuecomment-1783779687, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDKFWHZMLXQJZ5SW6ESLROLYBTRI3AVCNFSM6AAAAAA6PM5WGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBTG43TSNRYG4 . You are receiving this because you modified the open/close state.Message ID: @.***>

-- Dr. Peter Bailes P & C. M. Bailes Farming High Inhams Farm Stillington Road Sutton on the Forest York North Yorkshire YO61 1EQ Tel 01347-812589 Fax: 01347-811827 Mobile: 07796622921 Email: @.*** GPS 54 °04' 39.0" N 1°06' 11.3" W (54.0775°N,1.1030556 °W) We are what we repeatedly do. Excellence, then, is not an act, but a habit.

GregValiant commented 11 months ago

Lofts are always tricky for both the CAD software and the slicer to deal with. The math of a lofted surface can be quite intense. Since they are just gussets you may want to make them out of rectangles and then subtract a cylinder to create the radius.

The gussets don't appear to have been inset into the model. As a consequence when they were merged there is a gap between the vertical leg of the gussets and the wall of the model. It is always best to have overlap with (in this case) the gusset set into the model. The overlap would disappear during the merge operation. Trying to eyeball two surfaces to mate is tough unless the CAD package allows you to constrain the parts in all three axes.

In regards to the slicing - models like this that have a lot of geometric shapes on the build plate seem to be giving these last couple of versions of Cura some heartburn. This is the third model in the last couple of weeks to come up with slicing problems. (#17105 is similar) I'm going to re-open this so the Cura team will have another example of this behavior to use in their troubleshooting.

Yorkshire-Farmer commented 11 months ago

Dear Greg,

You are 100% correct. I have never tried printing a "Loft" before, but I found a tutorial on how to connect two different faces in a part design space and thought it worth a try - big mistake.

I agree the old fashioned approach to extrude and pocket is much simpler and safer. Version 5 of the gulley cover will be designed that way..

With regard to the closure of the bug report, if i did it , it was unintentional, I thought i was just uploading the project file to the open report, so no problem in re-opening it or using it for internal education etc.

Hopefully it will help others.

Best Wishes

Peter

On Mon, 30 Oct 2023 at 22:22, GregValiant @.***> wrote:

Lofts are always tricky for both the CAD software and the slicer to deal with. The math of a lofted surface can be quite intense. Since they are just gussets you may want to make them out of rectangles and then subtract a cylinder to create the radius.

The gussets don't appear to have been inset into the model. As a consequence when they were merged there is a gap between the vertical leg of the gussets and the wall of the model. It is always best to have overlap with (in this case) the gusset set into the model. The overlap would disappear during the merge operation. Trying to eyeball two surfaces to mate is tough unless the CAD package allows you to constrain the parts in all three axes.

In regards to the slicing - models like this that have a lot of geometric shapes on the build plate seem to be giving these last couple of versions of Cura some heartburn. This is the third model in the last couple of weeks to come up with slicing problems. I'm going to re-open this so the Cura team will have another example of this behavior to use in their troubleshooting.

— Reply to this email directly, view it on GitHub https://github.com/Ultimaker/Cura/issues/17103#issuecomment-1786138225, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDKFWH4EBXM73K47APITZULYCASEFAVCNFSM6AAAAAA6PM5WGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWGEZTQMRSGU . You are receiving this because you modified the open/close state.Message ID: @.***>

-- Dr. Peter Bailes P & C. M. Bailes Farming High Inhams Farm Stillington Road Sutton on the Forest York North Yorkshire YO61 1EQ Tel 01347-812589 Fax: 01347-811827 Mobile: 07796622921 Email: @.*** GPS 54 °04' 39.0" N 1°06' 11.3" W (54.0775°N,1.1030556 °W) We are what we repeatedly do. Excellence, then, is not an act, but a habit.