Fillet and boolean union problems

Next
 From:  Khris
8320.1 
Hi,

I am evaluating Moi 3D (trial version).

As a test I tried to model the main body a bathtub tap, which is very simple. But I have encountered several pb.

One I found a solution in the forum (separate + join on a curve). For the others I couldn't find the solution (I admit though I have not read all the search results :-) ).

There is only two parts on the model. One is simply a cylinder, the other the nose of the tap. I have made two versions of the nose. One (v1) by extruding a base shape. For the other (v2) I made a copy of the base shape, move it up, then selected the narrow end and moved it 5 mm towards the other end, and then used the loft tool between the two shapes.

I have attached the 3dm files to explain what I'm trying to do (and how I'm doing it), and to show the problems. I have named the files with a short description of the problems:

- "Pb fillet (unable to make fillets on red part) (nose v1)"
I want to make fillets on the whole red part. It works if I select only the intersection with the blue cylinder.
But it doesn't on the green edges (calculating never ends if radius is < 6 mm) nor on the magenta edges (doesn't make the fillets - only the contour it seems).

- “Pb boolean union” (nose v2)
There are 2 problems, and 2 files. I'm sorry I'm unable to remember the differences between the 2 files.
When I try to make a boolean union between the 2 parts, in one case MoI is unable to make the union (calculating never ends), in the other it removes the top of the cylinder.

I've tried on 2 different computers with the same results.

Please ask if something is not clear.

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
8320.2 In reply to 8320.1 
Hi Khris, thanks for posting your files. I took a look at "Pb boolean-union (top of cylinder disappears) (nose v2).3dm" and it looks like the problem there s the circular parts of the 2 solids are not quite precisely aligned, they are pretty close like around 0.002 units apart but that adds a lot of complexity to the boolean process when 2 objects have overlapping surface area.

Part of the boolean process involves intersecting surfaces of each solid with surfaces of the other solid, and when surfaces have overlapping area it adds more stages to the boolean process and it can be difficult for the intersector to decide if the surfaces are exactly overlapping or if they need some kind of intersection generated between them. So it can be good to avoid having overlapping surfaces or when you do it's good to have them more exactly aligned.

I also took a look at your base shapes file, and the issue is that these profiles come close to being aligned but are just off by enough to cause all kinds of intersection difficulties:






So that's likely to be the root of a lot of your problems in this case, to get a good boolean on those pieces you'd probably need to have those base shapes on the left side be exactly aligned with each other instead of what you've currently got. Either they need to be more closely aligned or less closely aligned, it's when they're about as close as you have them that will tend to get things tripped up because it's close to the intersection calculation tolerance.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
8320.3 In reply to 8320.1 
For your fillet case, unfortunately that's a lot more difficult kind of fillet than what you may initially think.

The problem is putting in fillets cuts away some areas of the object and would cause holes to open up, filling in the hole for this case is not handled well by MoI's current fillet engine:








Above I used Edit > Separate to break those 2 surfaces away from the main object, then with them as individual surfaces instead of joined together Fillet operates in a different mode called surface/surface fillet, it is able to generate fillet surfaces in situations where the edge based filleter can get confused. So it is possible to make some progress on getting those fillets done that way but you would need to do some manual work on figuring out how to deal with that hole area.

I'd recommend trying that kind of fillet case in another CAD program that has a more sophisticated fillet engine that handles more types of junctures, try OnShape or ViaCAD.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Khris
8320.4 
Thank you for this detailed answer, it was very informative !

Following your advice I've remade the 2nd shape so that it superimpose exactly on the 1st one on this side, and indeed the the fusion worked perfectly !

And I've found a way to make the fillet:

- Fillet the whole nose (1st fillet.jpg) (3 mm)
- Make the union
- Select the intersection with the cylinder (2nd fillet (selection).jpg) and fillet that (3 mm)

And now I have a nice clean fillet !

Thanks again !

Image Attachments:
Size: 281.9 KB, Downloaded: 21 times, Dimensions: 1426x872px
Size: 279.1 KB, Downloaded: 18 times, Dimensions: 1426x872px
Size: 238.6 KB, Downloaded: 24 times, Dimensions: 1426x872px
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  Michael Gibson
8320.5 In reply to 8320.4 
You're welcome Khris, I'm glad you were able to get it finished! Yes, doing the fillet beforehand in this case is a good idea since it avoids making a difficult end condition for the filleter to handle.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All