Fillets and old iron

[drmacro]

I mentioned in a post the other day that I was working through exercises in an old drafting text.

I saw this one and thought it would give filleting a hard time...I was right.

I've attempted this in Part Design only, mixed Part/Part Design (i.e. think base features), and Part only.
I've been unable to get all the fillets.

Fig_6-91.jpg (516.18 KiB) Viewed 345 times

I'm willing to be proven wrong.

OS: Ubuntu 19.10 (XFCE/ubuntustudio)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.
Build type: Release
Python version: 3.7.5
Qt version: 5.12.4
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: English/United States (en_US)

[carlopav]

I think much of the problem with fillets it's here:

Cattura.JPG (25.71 KiB) Viewed 328 times

You can set different fillets for every edge, so you have enough surface to be filleted.
EDIT: also you can treat every fillet at the end and not include some of them in the sketches.

[drmacro]

I think much of the problem with fillets it's here:
Cattura.JPG
You can set different fillets for every edge, so you have enough surface to be filleted.
EDIT: also you can treat every fillet at the end and not include some of them in the sketches.

Everything you said.

But, the reason that I attempted to do it along the way (in that attempt) was, doing the fillets at the end literally crashed FC.

In addition, different fillet radii can be used...but, would not complete the exercise as prescribed. I.e. the note: "FILLETS & ROUNDS 1/8 R"

If you look close, I even fudged the area you highlight by making the extrusion extend past the flange of the revolve. To avoid sliver faces where the flange revolve meets the pad.

[bejant]

When that textbook was written the mold pattern would have been made by hand from wood and the outside corner fillets abraded away (file, sandpaper) and the inside corner fillets made by filing them in with a bead of some material that hardened afterwards.

I don't think that directly trying to model the piece in the as-machined condition can produce a proper model. Instead, I think that the un-machined casting should be modeled, and once that raw casting is complete, the machined holes and flats should be pocketed away.

Even so, I don't think the 2D drawing is a true representation of the 3D part, especially at the back where the fillets on the vertical side faces meet the curve flange thickness and at the front where the machined face meets the flange corner fillets. (Quick61 had a word or phrase for that type of visual hocus-pocus but I don't remember what it was).

[tanderson69]

opencascade can do it with the proper amount of swearing. I had the most trouble in the front area without the boss.

[bejant]

Wow ! Looks nicely done - can you post the file?

[tanderson69]

Wow ! Looks nicely done - can you post the file?

I used opencascade, but not freecad. So all I can give you is a brep. If you want to know how I did something specific, just ask.

[carlopav]

I used opencascade, but not freecad. So all I can give you is a brep. If you want to know how I did something specific, just ask.

What does I mean exactly? Sorry but I'm very curious...

[user1234]

Just for testing. Works without problems here (PartDesign). But i was expecting errors on the corner (like other CADs) so i the way i model it was a little different (like i do when i see corners like that). You can model it much easier (treestructure) but for me it is ok. I rounded all measures in metric.

Greetings

info:

Code: Select all

OS: Debian GNU/Linux 10 (buster) (X-Cinnamon/lightdm-xsession)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.19524 (Git)
Build type: Release
Branch: master
Hash: 045e924cd6ef35f03c59ab5dbd07ad9e78508482
Python version: 3.7.3
Qt version: 5.11.3
Coin version: 4.0.0a
OCC version: 7.4.0
Locale: English/UnitedStates (en_US)

[carlopav]

Great! this would be a nice tutorial btw