printing ball issues

[udir 0]

Hello

I am trying to print a mold for a flatted ball (see "half mold" image). The resolution of the model is high (about 1M points, the STL file is about 100Mb - too heavy for sending...).

Cube areas are sliced and printed perfect.

Flatted ball areas are sliced strange and printed badly (see "sliced image" and "flatted ball...").

On printing a ball, I have faced similar issues.

Are there any special parameters / method for slicing a ball?

Any idea what I am missing?

Thanks a lot

[Longtoke 25]

Very odd....

 

I tested the issue by creating a basic sphere ( 80mm diameter ) and using a basic cube base of 100mm x 100mm x 50mm to losely replicate your model, but a very low res version Tremendous Wluff.stl ( made with tinkercad ). 

 

Did the slice on a 0.2 layer height and the same odd pattern appears in the low point center ?!?!

 

It looks like Cura is getting confused by the shallow gradient of the curve at the bottom ( note the tell-tale concentric walls in the preview ).

 

Now really weirdly, even with the concentric walls, the layering still seems correct ( from the Cura preview ) on the low res poor mans version I did anyway. From the look of  your model failure, I'd suggest having a few layers of material below the lowest point of the sphere, as it looks as though you've  had adhesion issues due to the mix of small concentric walls and diagonal fill. Even a 2mm layer below it could be enough to sort it. 

 

Good luck and let us know how you go. 

[GregValiant 1,454]

Large radius objects are tough.  The angle is severe at the lower layers causing the printer paths to be far apart and they don't want to bond.

I think I'm in agreement with @Longtoke regarding moving the ball up off the build plate (2mm sounds good) and letting the underside be all support structure.  I know it won't leave a great finish, but it should print.  Another thing to try would be Adaptive Layers.  

[Jaka3d 1]

My advice:
Rotate model for 90° in X or Y axis and you'll get better results.

In your situation, i would rotate model for 90° (Left surface (on your image) align to the build plate).

 

[udir 0]

13 hours ago, Longtoke said:

Very odd....

 

I tested the issue by creating a basic sphere ( 80mm diameter ) and using a basic cube base of 100mm x 100mm x 50mm to losely replicate your model, but a very low res version Tremendous Wluff.stl ( made with tinkercad ). 

 

Did the slice on a 0.2 layer height and the same odd pattern appears in the low point center ?!?!

 

It looks like Cura is getting confused by the shallow gradient of the curve at the bottom ( note the tell-tale concentric walls in the preview ).

 

Now really weirdly, even with the concentric walls, the layering still seems correct ( from the Cura preview ) on the low res poor mans version I did anyway. From the look of  your model failure, I'd suggest having a few layers of material below the lowest point of the sphere, as it looks as though you've  had adhesion issues due to the mix of small concentric walls and diagonal fill. Even a 2mm layer below it could be enough to sort it. 

 

Good luck and let us know how you go. 

Layer height is 0.2mm, the odd pattern appears in the low point center. I face the same issue when printing a ball …

[udir 0]

11 hours ago, GregValiant said:

Large radius objects are tough.  The angle is severe at the lower layers causing the printer paths to be far apart and they don't want to bond.

I think I'm in agreement with @Longtoke regarding moving the ball up off the build plate (2mm sounds good) and letting the underside be all support structure.  I know it won't leave a great finish, but it should print.  Another thing to try would be Adaptive Layers.  

I have already used adaptive layer in the described issue (max variation 0.2, variation step size 0.1, topography size 0.1)

[udir 0]

6 hours ago, Jaka3d said:

My advice:
Rotate model for 90° in X or Y axis and you'll get better results.

In your situation, i would rotate model for 90° (Left surface (on your image) align to the build plate).

 

Sounds a simple change and good idea. I'll try it, but it couldn't be a work around for printing a ball …

[Jaka3d 1]

On 10/23/2020 at 3:37 PM, udir said:

Sounds a simple change and good idea. I'll try it, but it couldn't be a work around for printing a ball …

Once, when i wanted to print a ball, i cut the 3d model to 4 pieces (quarters) and print it. I rotated every piece for 45° i got very pretty outer surface (need of support structure due to small conntact area to the buildplate). Then I glued all pieces together (use holes/pins/slots) and it was very nice resoult, prepaired to postprocessing.