blood cells in a blood vessel

[Haynes]

Hey guys,

I'm currently working on a 'blood cell flow' test in ICE and so far I'm content with the cells' movement except for the fact that they are still interpenetrating.
Since the particles' density is supposed to be quite high and their size (in the picture) not insignificant, one can clearly see every single penetration, so I was wondering if there is a 'simple' method to prevent the particles from piercing each other. As far as I've heard and read there is no 'simple' solution for that issue, but I hope there's is at least a way, simple or not, to keep the particles 'at bay'.

In order to prevent the blood cells to penetrate the vessel's surface I used the 'surface force' node, which worked decently, even though not 100 % accurately. So I guess there must be a similar method for the particles, although I understand that this will be a way more complex simulation.

I figured out so far that the node 'get closest location' should be helpful but since I'm an noob concerning complex ICE stuff I was hoping you could give me a hint how to solve this issue.

Best regards and thank you very much in advance

[kiwimation]

Can you not just use bullet Rigid Bodies in ICE?

[luchifer]

In advance, im sorry is this sound amateur but, why not use a low poly mesh as a proxy ? (Only for collisions) That way you shoulndt calculate too much data

[Bullit]

Are you using rigid bodies?

[Haynes]

Thank you, guys, for your prompt response.

@luchifer
you are right, for a faster calculation I will definitely use a low poly mesh for the blood cells. I just used the high poly mesh to see if my computer gets along with it:)

@Bullit (& kiwimation)
No, not yet! I'm using a simple 'Flow along Curve' for the movement and a 'Surface Force' that prevents the cells from touching the vessel's surface.

Actually, using 'rigid bodies' was my first idea, but then I decided against doing so because I thought that there's no real 'bouncing off' effect, since the blood cells flow within a liquid and don't really collide. At least this is what I have in mind, when I think of a blood flow. But I might be wrong and the rigid bodies method works perfectly. I will give it try...

But nevertheless: do you think that there's way (apart from the rigid bodies) that makes the cells repel each other, something like the 'surface force'?

Best regards

[Bullit]

I can only think of some repulsive force like negating a coagulate force.

[Haynes]

Yeah, using a simple repulsive force that affects the particles sounds reasonable and this is what I had in mind right from the start but apparently this effect is not so easy to achieve when you don't want to use rigid bodies:)
Thanx & best regards

[Bullit]

Were you able to fix it?

i have got another idea. Every cell has a bunch of very small cells attached to it, but are hidden, that might be possible to push the others out a bit more.

[Haynes]

Hey Bullit,
not yet, really! This task proved to be more complicated than I thought:) Thank you very much for your input, I`ll keep that in mind.
Best regards

[FXDude]

Bullit suggested a negative coagulate (to repel each other on the particles, on top of your surface force of your tube, did you try that?

Also, if you still want to try dynamics, there is sample scene showing soft bodies that has a circus with a hammer that bumps like an inflated clown down into what looks like one of those really old cone-shaped record player speakers. (dont recall the name of the scene)

Once I tried duplicating the rubber clown, and assinging the other clown as an additional obstacle on each of them
(also checkboxing "Animated" in obstacle settings)
putting both clowns on top of the cone/tube and then simply letting them drop drop in,
and they would both bounce-off the cone and eachother pretty well, (no artifacts of funny stuff) before going into the center tube.
(while not exactly lowres models and set to "exact shapes" it ran at some ~24 fps)
and I said huh!

Aslo Soft bodies even if set to somewhat solid look settings may yeild more organic feel than simply changing directions when they touch,
or you can set them to be somewhat "smooshy" for them to like bounce slowly. (extra bounding geo can also be used here)

So how about dropping cells in a tube with softbodies, assinging each other cell as obstacles on each + the tube?

I'm sure ICE syflex (or lagoa) could also do it, but never tried setting that up. (yet :] )

[Haynes]

Thank you very much for your response, FXDude!
No, so far I've haven't used the coagulate force but I'll give it a try as soon as possible. And I'm gonna take a look at the sample scene you mentioned. That method sounds really promising!
Thanx again & best regards