Cineversity

So here I am again having to animate/simulate several hoses and a power cord for a product that rolls over the ground, pulling a hose and power cord behind it.  The idea is to start with the hose coiled up, then uncoiling as the product moves away, pulling along the hose.  I’m getting pretty close starting with a modified helix, adding Spline Dynamics and constraining one end to a faucet and the other to the product.  I have a collider for the ground, and that works fine.  The problem I’m having is that when the helix coil settles it intersects itself on the ground collider.  Is there any way to make my hose self-collide so it doesn’t cut through itself?  It seems like even if I model the coiled section so it’s not intersecting at the start, once I run the simulation gravity will pull the spline coil down to the ground collider, intersecting itself again.  In a C4DCafe forum someone talks about turning off the gravity for the section of coiled rope using a gravity force object with inverted falloff.  I might try to figure out how that would work.

Having the simulation figure out the self-collision for a 3 cm diameter hose would be ideal, though.

Thanks.

Shawn Marshall
Marshall Arts Motion Graphics

Hi Shawn,

I would suggest two ways, though—I haven’t used/played with those in a while. I hope I do not sound de-motivational, but I haven’t found something that I would call production-ready, if based on Spline Dynamics. Perhaps I expect too much from it or I miss something. ;o)

1) Create a chain with MoGraph Dynamics and calculate the chain. (I have attached an image, think of setting it to Shape>Moving Mesh in the Collider Tab. From my point of view this is overkill and not the way I work. However, sometimes it might be worth a shot, even if I think it takes a lot of patience to get convincing results. The chain-elements need to be used then in an MoGraph tracer to create the base spline for the hose. A Friction field might help to calm down the chain. Again, not the way I like to work, but it simulates the distance and chain behavior to a certain degree.

2) Get the hose as far as you like to have it, then create a PLA from it and adjust the few areas where it over laps manually.  If there are too many, split the hose perhaps in part s that are on the floor and the ones that are on the second layer, adjust these or place for the second layer extra (invisible) colliders.

Yes, it would be nice to use Hair with thickness for that. I guess a complete hose needs a lot of calculation to get it right. I tend to animate manually then. That sounds maybe sloppy for some people, but I like control and as long as we don’t have a fast and easy way to it, I stick with it. Perhaps I miss something, but think of the split (separate the spline for the hose) and perhaps a deformer might help.

A Displacer Deformer with a map (image) perhaps. Just displacement where the first segment of the hose is on the ground, you can even render this.

Perhaps you might make a suggestion to MAXON. http://www.maxon.net/support/suggestions.html

All the best

Sassi

Hi:

Thanks for taking the time to reply.  I ended up using a spline dynamics simulation on some of the elements and hand-keyframed another element.  It didn’t look great but will suffice for this first draft I had to get out before our vacation.  I’ll do more tests and R&D when we get back.

Take care.

Shawn

Hi Shawn,

Yes, sometimes there is manual work the only way—even if it is so easy to define what is needed in terms of functionality. The dynamic spline is certainly an area where I found, based on forums questions, some room for improvement. To suggest that to MAXON better more often than not will help to push those features more toward the top position of the to-do list. If no one ask for it, other features with more artist-request will win.

Have a great vacation. I will keep it in mind…

Sassi

P.S.: As I said I will keep it in mind. In the past days I tried several “ideas”, but it is tricky, to say the least. I have attached a “sketch” that I consider the most promising, even if I’m not certain if that is really a solution. take it as “brainstorming” or “WIP” (work in progress).

The idea is to use a MoSpline with the initial Spline/Dynamic, and create an dynamic-collider. To make that work I pulled the resulting “Tube/MoSpline” a little bit up. This prevents a conflict or a collider-feedback-loop.

The image below shows the “dynamic-collider”, not the final hose of course. The final hose should be based on the original Spline, at least that is my initial idea about.

Have a look, perhaps it gives you an idea how to build up on it. My best wishes :o)