shrink wrap

[super gnu]

hi again!  trying to achieve a plastic shrink wrap effect.     

I was originally intending to do it in mcloth, but it does not allow negative pressure. 

fortunately, tyflow arrived just at the right moment, and *appears* to allow negative pressure for deflating stuff. 

however try as i might, i cannot seem to get any kind of sensible result.    im obviously missing something. 

ive tried up to 1/32 timestep and up to 30 solver steps..    ive tried all sorts of combinations of collision radii, and any number of different pressure settings in the "modify bindings" operator.  

all i get is basically a mess!

even positive pressure doesnt seem to do what id expect.   suggests user error. 



im attaching a simple test file (2018).. hopefully somebody can explain what im missing.

[peshang]

not sure if this is what you're looking for .. but i have got this result:

[super gnu]

well thats what i was hoping for from this test.. obviously in a real scene id model the "shrink wrap" to fit the target properly. what was i doing wrong?

[tyFlow]

Thanks for your scene file, I'll take a look at it.

Although negative inflate values don't really translate to negative pressure the way you might intuit...I mean, it equates to some form of pressure from the outside of the cloth, but not the type of vacuum inside the cloth that shrink wrap would need.

For a shrink wrap effect, what you actually want to modify is the "bias" value of your cloth bindings. A value less than 1 means that after initialization, the solver will think that all of the bindings should be shorter than they are, and will effectively pull them all inwards to compensate.

So if your cloth is composed of bindings that are 1.0 units in length and you set bias to 0.5, the solver will think the cloth has been stretch x2 and will shrink down. Same goes for the inverse....set bias to 2.0 and the cloth will blow up (without internal forces, just expand as it thinks it has shrunk down).

So proper bias settings are important for shrinking/expanding effects. Good for garments too, if you want to simulate tight seams.

[peshang]

my apologies for the slow reply,



not really sure what was wrong with the scene file that you had and i recreated the simulation from scratch ..

cant tell if it's a unit system issue ..

[MeHoo]

I've done this in max cloth numerous times when I was doing consumer packaged goods modeling.  Stuff like hot dogs, bacon, etc.  https://drive.google.com/open?id=11Y05ny...o7y_cCjMFc

I'll have to give tyFlow a try to see how it operates, but like Tyson said, negative pressure isn't really negative pressure, and sometimes I'd actually get better results inverting the normals and applying more pressure if that makes any sense.. found that out on accident after a hot dog sim ate itself lol.  https://drive.google.com/open?id=1-84cWk...M8McJPucuM

I was playing around in the file and noticed that the timing on the modify binding was not set to continuous which means, if I'm not mistaken, that it is only operating once and not re-iterating over and over, which is how a real shrink-wrap works.  Changing that gets something more like a shrink wrap.  Is that what you were looking for?  In either case, it looks cool. 

   

The bias like Tyson said is like a rubber effect where the material is able to shear or shrink on itself.  In this example it would probably lessen the wrinkles that get caught if the self-friction doesn't hold them together.   Most of my shrink wrap work was with a more plastic material.

I also reduced the collision particle size because once the negative bias takes over they seem to shear weird.  This seems to look good, albeit it's a little slower.

Here it is with some negative bias:     
 I'll attach that file next.  Hope it helps.

---

negative bias file: shrinkwrap04.max

[emmasla]

Hi! 

I have been struggling with shrink wrapping a tube chair I made for a while, and then found this discussion. 
The negative bias shrink wrap MeHoo did is exactly what I need, so I downloaded the shrinkwrap04.max file and merged it with my file.

Due to the complexity of the chair I didn't manage to get a decent result. I modeled a birth object that's similar to the shape of the chair, but it's still not properly colliding with the chair and the cloth seems to be shredding. (see attachments)

I've tried adjusting some values such as the multiplier, the collision offset, and the inflation force - which slightly improved the wrap on some parts, but as you can see in collision-2.png it's still not at all working. 

Which values should I adjust to solve this?

Thanks a lot!