Variable fillet not copied in circular pattern

[andrew.harpur.whc]

I'm creating an impeller example and have generated a single body comprising of the hub and a single blade.  I've created a variable radius fillet between the blade and the hub.  When copying the blade and fillet as a circular pattern, only the blade faces are copied and the fillet is lost - (Solid>Create>Pattern>Circular Pattern)

It's not acceptable to recreate the fillet on every blade. See images and model attached...

[jeff_strater]

my initial look at it seems to be that the geometry of the variable fillet is blocking it.  The fact that the fillet just touches the outer edge at this point, I think, is the problem:

If I change the radius at this point to 7mm instead of 8mm:

the pattern can work:

I don't know if 7mm fits your design requirements or not, but that is the way to fix it.  I did not try other values.  Model attached

[TrippyLighting · Accepted solution]

Or you offset the lower rim +5mm , then create the feature pattern and then create a "New" (not "Automatic") offset -5mm.

[g-andresen]

Hi,

Although I can't verify it at this time, I would try to use faces instead of features to create the circular pattern.

Günther

[andrew.harpur.whc]

@jeff_strater : Hi Jeff,  Thanks very much for taking time to look at this. 

I did try changing the 8mm part of the variable fillet as you suggested, but unfortunately it didn't work for me.  Eventually I downloaded your own file (Impeller Blade (js).f3d).  Looking at your own file; yes the fillet has been copied with the blade, but the variable fillet has changed shape as can be seen in your 3rd image.

I played around further with my original model and managed to get the blades and fillet to copy around the axis for the majority of blades - but then the original fillet is still lost in one iteration and the copies are distorted at the top of the component...    new image and model attached.

I think there is a real issue with the variable fillet here that needs logging as a bug with the support team.  It's an example project and not time or geometrical critical.

Once again, thanks for you time.

[andrew.harpur.whc]

@TrippyLighting : Thanks for your time Peter. That hack provided a good work-around. 👍 👏 😀

[andrew.harpur.whc]

@g-andresen :  Thanks for your input Guenther. I did try both faces and features at the time.

[davebYYPCU]

Original file - no bug, just a workflow thing. 

Changed the press pull to additional Revolve. 

Combine to the hub is required for the fillet, but is not helping the pattern. 

Make a copy body of the hub  and

edit sketch 1, adding a top section to the sketch as an additional profile - for later.

After the blade is filleted, revolve cut the hub (body 1) off the blade.

pattern the blades, unhide the 2nd hub, and then make the (revolve) small top hub again, as Join - joins it all together.

Might help.....

[TrippyLighting]

---

@davebYYPCU wrote:

Original file - no bug, just a workflow thing. 

I would disagree with that assessment simply because it is too dismissive for my taste.

We can label this whatever we want, but it definitely is an area with room for improvement and should be recorded as as such.

[davebYYPCU]

Original statement is patterning blade with the hub.  

My workflow - no errors.