An experiment with a better collision model for hair.
Igna Valer writes:
I did some simulations of the hair and I found a solution of hair collision. I think it`s ok but have some problems, mabe someone can tell me how to make better hair collision.
I post on youtube the simulations and the settings of the objects.
Tests:
Explanations of the hair collision:
20 Comments
Good workaround until we get real hair collision.
For speed problem , probably increasing substep works.
Hope to see "hair collision" patch integrated into the trunk or any better
solution. Workaround is ok but not the real thing.
Well.. ok, good job, but I don't think it's pleasure to work with this kind of workarounds all the time. I am waiting for hair collisions about a year now.
+1 with Karlis
Have you tried changing the cache step in your hair particle setting to 1 instead of the default 10? It might help with the scenario where you try to move the colliding object really fast.
I played with this a bit more. You can get the hair to collide with a fast moving object if you do the following. Keep the 'force field' properties, but also enable 'collision physics' to the object that the hair is colliding with (The default values are fine). Next instead of a strength of 100 or 200 in the force field settings, use just 1 (all other settings you mentioned in the video remain). Finally, in the particle settings for your hair emitter set the cache step to 1 under the cache panel, and under 'field weights' at the bottom of the particle settings, type a value of '50' as the 'Force' value. The slider only goes to '1', but you can manually type a higher value in. With these settings I was able to achieve 'high speed collisions'. Hope that helps.
Thank you :)
I have a theory about forcefield moving from bottom.
Maybe endpoint of hair movement are determined by sum of 3 vectors. Hair weight vector, some kind of spring vector (that resist gravity and don't allow hairs to fall), and force field vector. When forcefield is is applied from the bottom, sum of 3 vectors are equal to zero. Because weight+spring vectors is parallel and opposite direction to forcefield vector. But why that opposite direction vectors length is equal?
Warning: my theory is not based on any experimental or theoretical data, so probably incorrect.
Does Igna Valer have an account on BlenderArtist for which to contact him, or any other prefered method of contact?
Hi,
Sorry i don`t have a acount of BlenderArtist but i can make one. If it`s ok i let you my e-mail adres: [email protected] or [email protected].
[email protected]
You might want to check out the patch that Helmut Schultz created. It adds Blender's existing cloth collision to particle hair.
Thread on Blender Artists: http://blenderartists.org/forum/showthread.php?222788-Collision-hair-particles
THANK YOU :) Niverik2k have send me one e-mail with this solutin and it`s seems to be the right solution for my project. :)
I have an idea for solving the gap problem in the collision: create 2 objects, one within the other, with the force field attached to the inner one. The render will only show the outer object, while the collision is actually calculated against the inner one. With the right relative sizes of the two, the gap around the inner object should exactly match the size of the outer one.
Anyway, worth a try, I guess.
not to toot my own horn, but i totally posted this exact technique on blenderartists almost a year ago and no one seemed to notice!
http://blenderartists.org/forum/showthread.php?208672-hair-collisions-in-blender-2.5&p=1785524&highlight=#post1785524
ok, so thats pretty much the definition of tooting my own horn, whatever :P
WOW i havent seen your post, if i would, i didnț spent my time trying to find the proper simulation. Maybe in the future Blender Nation or Blender could gather all the informations for us.
no worries, i was just being obnoxious for the sake of it ;)
its neat that we both arrived at the same solution though, and you've provided a much more thorough and better demonstrated explanation in your videos!
Thank you all who help me to find the right solution for hair collision.
I don't know if you already found a solution to this but I agree with Zomby that it probably doesn't work with a perfectly perpendicular angle. Maybe rotating the forcefield as little as 0.01 degrees might fix it.
Looks good, but what's going on with the black crawling colors all over the scalp in that test video? What's going on there?