Can I place down points on specific spots for variable fillet

[1297193504]

I have these two tubes joining at an angle, and I want to give the seam a variable fillet. So how can I place a point on the exact top/bottom/sides. I can type in 0/.25/.5/75 but because the tubes are at an angle, these numbers don't exactly match the points I wanted. Is there a way around this? I tried Inventor and had the same results.

[davebYYPCU]

You just click on the fillet edge, you can type the position into the dialogue box.  Percentage of the length of the fillet curve, so your points are 0.25, .05, .75, 0, etc, with radii values that are going to work.

Might help.... 

[1297193504]

It seems that you have your tubes perpendicular to each other. The problem I'm getting here is that if they are at an angle other than 90, the numbers 0/.25/.5/.75 won't work, these points won't be on he spots that we suppose them to be on.

[davebYYPCU]

Nope.

What angle would you like me to demo?

Lower than this needs radius adjusted to something possible.

Might help....

[1297193504]

For what I've tested, when the angle is not 90, the points 0 and .5 aren't on the exact opposite of each other. Any clue what happened?

[laughingcreek · Accepted solution]

splitting the face of the pipe will create snap points for the variable fillet. see attached model. note these are snaps and not constraints. you will have to repick the points if you adjust the angle.

[TrippyLighting]

This looks like the joint between a head- and top tube of a bicycle frame. Is that a correct assessment?

If it is I would create lofts and stay away from solid modeling and variable fillets.

More work, often fiddly but much better control over the shape of the transition. 

[1297193504]

that seems to be a good idea, but is there a command to get rid of the seam line after filleting?

[1297193504]

You're correct. I've done another model using surface modelling and the process was a real pain in the ass. For this one I am making the lug structures, and the casted steel ones sold online looks like they just have a simple fillet.

[TrippyLighting]

---

@1297193504 wrote:
... I've done another model using surface modelling and the process was a real pain i

---

If you want to share the model I'll look it over. Perhaps there might be improvements possible ...

If a variable fillet would not be a PITA, you would not have come here 😉

[TrippyLighting]

---

@1297193504 wrote:
that seems to be a good idea, but is there a command to get rid of the seam line after filleting?

---

No! NURBS have seams, that is just how they work in CAD software. Some CAD software packages don't show the seam, but there is always a seam.

[1297193504]

It is still a wip. Surface modelling is just way more complicated, and even making parametric changes could lead to broken models. The last one I did took me dozens of hours of work, I'm not ready for another one.

[laughingcreek]

---

@1297193504 wrote:
... is there a command to get rid of the seam line after filleting?

---

Mostly the answer will be to leave it.  it doesn't effect the surface quality,  and doesn't show in renders or in drawings.  

but if you really find it so offensive that you've got to get rid of it, one approach that SOMETIMES works is to select one of the faces and DELETE. (while in the solid workspace, not the surface workspace)

before-

after-

You have to careful with this approach because it sometimes changes the intended geometry of the model.  but not in this case.

another approach to getting rid of it is to manage the extent of the split so it doesn't run the length of the part. (this is the approach I usually employ incases where the split will interfere with operations on the other end of the part)