Full Version: Shell Options & 3D Printing

I ran into an issue that, at first, was baffling to me. I had created a revolved shape, run the SHELL commmand and then cut holes in the shelled object. No paying close attention, everything seemed OK until I went to print the object. Bamboo Studio then required me to perform 'Repairing the Non-Manifold Edges" When the repair was complete, all the wholes had been completely filled in.

In chasing down the problem, I used InStep Studio and was puzzled by its claim that my STL was actually TWO objects when one was expected.

It took me a while to realize that the default ACIS option for shell was creating an inner and outer shell. Using the Solid++ option did as I had expected. The inner and outer surfaces were part of a SINGLE solid object.

I wonder why I had never noticed that before now?

Tom

Image Attachments:
Shell Options.png 

---

I could probably point out a couple bugs in ACIS shell also if you are digging into that part and want/ready to hear.....

Just reproducing Toms object with a sphere can lead into a couple that may point to something....

Create a sphere of radius 12. Stretch it down with edit frame to look like Tomes object. Shell it 1 inch. Try to cut it in half with a Boolean diff line front plan. Spin 180 same result. Spin 90 cuts.

Delete outer shell with a drill down selection before spinning. cut solid in half. Attempt 1 inch shell on half.

---

Hi Tom, can you please send me the .3dm model file with those objects?

- Michael

---

Sure. :)

By the way, It is amazing how many new capabilities you always manage to bring us! Thank You!

Tom

---

Here is a sample file. It shows Solid, Shell ACIS & Shell Solid++ and includes cut versions of the shelled samples.

NOTE: I had not seen the artifacts produced by cutting the Solid++ version until preparing this file for you. Up to this point the cuts were performed cleanly. I left them in so that you could assess what might be happening.

Attachments:
Sample Shells.3dm

---

Thanks Tom, it looks like there are a couple of tuneups needed for conversion to/from ACIS geometry when there are internal void regions.

I'm investigating.

- Michael

---

Thanks! I'm happy to have been a part of seeing the issue identified before final release. :)

---