Thank you for testing!

I was able to fillet the spokes against the inner hub as you did, but the outer fillets failed 4 out of 5 times. So very strange, since they are an array of the spoke that filleted correctly.

Cheers,

Chris

Hi twofoot,

after testing Pilou's suggestion i came up with this to get all elements filleted.

I cut your model in five pieces and then radial arrayed one filleted element, to obtain a solid.



Regards
Stefan

""""""""""""""""""So here's the file without any fillets. Does anyone have suggestions on how to add 6-8mm fillets on the yellow highlighted areas for each spoke? The multiple tapered sections seem to confuse the fillet engine, and what works for one spoke *does not work for all*."""""""""""""""""""

There are a few things with this file I didn't like. Some surfaces were not smooth? (Separate outer surface of inner hub and zoom in)

Mostly though the fillets failed because of a slightly twisted, out of plane surface structure which made for terrible fillets. Here is a video where I rebuilt objects.

at beginning I planar (flat snap) spoke objects. Fillet then works ok at 8mm.

Success! Thank you for this rather clever idea Stefan.

Cheers,

Chris

Burrman thanks for putting this together.

The file is going to be 3D printed for use as a foundry pattern to cast these wheels in ductile iron. The taper (1.5 degrees) needs to be there to allow the pattern to pull cleanly out of the molding sand at the foundry.

Chris.

Hi Chris,
I noticed tapers on the hub and the inner wall of the outside rim and understand a draft for casting. However, i did not "remove a taper".....

I did view the spokes as a "straight extrude in z" and thats what i did in the video, but thats not the issue i demonstrated in the video. I will re-examine the spokes draft angle and redo the vid if necessary. The issue is with regard to non planar curves being used to generate the solids.

Sorry Burrman. I blame it on looking at this on my mobile. ;)

C.

I double checked the original files spokes and there was no draft set there. Here is the file I was working on in the video.

Hey Chris,
Just made another video to show the particular area I was talking about.



Notice in the video that regular 2d "planar curves" will only have an X+Y value? so a curve that is non planar will show a Z value also (Its a very small value)

I have a program which will display out to 15 decimals and the Z number on those curves is "0.000065326690674"........

That number changes as you set decimals with rounding. So even the others you saw that said "0" in MoI, were out by some factor.

I think this is what was messing with the fillet calcs (or better said MoI's display mesher with bad trims) and why the STl still calced out. (Number that was out is VERY small)

Anyway, maybe look into why you got non-planar curves will help you to avoid this in the future...