H20 - feathers - easy way to make the shaft stiff?

   997   6   2
User Avatar
Member
34 posts
Joined: March 2023
Offline
I am simulating a single feather in vellum.

The entire workflow works fine.

Only the shaft is even more limp than the barbs when simulating, no matter how high
I set the stiffnesses of stretch and bend.

Any ideas on how to make the shaft stiff in a vellum sim?

What am I missing?

Thanks so much!
User Avatar
Member
7771 posts
Joined: Sept. 2011
Offline
I know vellum doesn't like too many divisions or stiffness becomes impossible to enforce. The only way to increase stiffness is to increase iterations or reduce subdivisions.
User Avatar
Member
34 posts
Joined: March 2023
Offline
Thanks! I know that.

I was specifically referring to the feather workflow.
And treating the barbs differently than the shaft.
Which just makes sense.
User Avatar
Staff
637 posts
Joined: Aug. 2013
Offline
You probably have something like 500+ points on the shaft curve if you're trying to simulate the full feather. There's not much of a way around that because you need the barbs connected.

But you will always get a floppy curve with that many points.

As a workaround, you could copy the shaft curve and resample that to 8-20 points, set up constraints on that and also constrain it to the real shaft curve. This could stabilize the whole feather and require fewer constraint iterations. I haven't tested this though. Make sure collisions between those overlapping curves aren't computed.
Kai Stavginski
Senior Technical Director
SideFX
User Avatar
Member
34 posts
Joined: March 2023
Offline
KaiStavginski
You probably have something like 500+ points on the shaft curve if you're trying to simulate the full feather. There's not much of a way around that because you need the barbs connected.

But you will always get a floppy curve with that many points.

As a workaround, you could copy the shaft curve and resample that to 8-20 points, set up constraints on that and also constrain it to the real shaft curve. This could stabilize the whole feather and require fewer constraint iterations. I haven't tested this though. Make sure collisions between those overlapping curves aren't computed.

Hello Kai!

Thanks for your quick answer.
That sounds like something I could do.
But you get the idea?

Maybe something for an update?

At least a shaft group output, maybe on the feather interpolate?

Best regards!
User Avatar
Member
617 posts
Joined: Aug. 2008
Offline
sim a curve with less and then you can wiredeform the one with the feather.
User Avatar
Member
34 posts
Joined: March 2023
Offline
pelos
sim a curve with less and then you can wiredeform the one with the feather.
I'm sorry, I didn't understand
that.

Would be happy if you could be more specific.

Thanks so much!
  • Quick Links