ChargingImage

I wonder... nobody uses those constrains? Everyone does it on other pins and transforms them "internally"?

Hallo! I want to somehow delete pins that are holding cloth, but have several issues with this. In simple example, pins get "ungrouped" by solver, but everything is disconnecting (there should be green balls). cloth_constraint_delete.hip I made very simple example file (that also shows how boolcut is destructive to cloth). My main problem is from more complex scene: What I couldn't replicate in simple example, but this is what happens when I try to add sopsolver (same as in example file), that is just removing points from group. Without it - it works fine. Points are not deleted, reordered, does not change topology. Last thing is WHY I'm trying to delete those pins: This is what happens when pins from similar spot are being separated by rigid simulation (rigid bodies hidden). I'm suspecting that it's when one "cloth poly" is attached to two different rigid objects. It won't tear, it will just stretch. Any ideas?

I'm mainly thinking about simple bending, with possible ripping. From I remember I had trouble with rest position not updating correctly. It was unstable or just wrong. But melting approach would be cool too. For the future.

Does anyone did successful metal bending FEM test? I've tried what is around in forum, but without luck...

Hey Atom! Unfortunately I've read that like 10 times. Not sure if I understood all correctly, but I know the topic. It kind of helped me with understanding problem, but not solving it. Most simple cry out for help would be: how to force smoke to behave like simple, "lifeless", unturbulent, dull, boring particles. Or maybe, how to render boring particles as nice looking smoke (no luck with that either).

No one can help? Maybe I wrote it a little bit too confusing...

Hallo! I'm trying to figure out how to properly do space explosion / impact on mars or any other planet that has no atmosphere. I've read huge topic about smoke behavior: http://forums.odforce.net/topic/18397-get-rid-of-mushroom-effect-in-explosion-and-add-details-in-the-opening/ So let's begin from simple things. REFERENCE: so behavior of dust/particles is... super simple. About like this: Ans this behavior is everything I can't achieve with smokesolver. Maybe I'm idiot. Test of explosion: 1. microsolver "projectmultigrid": When I turn off this microsolver it actually does what I'm trying to achive: no mushrooms because of no pressure. Thing is, I'm trying to somehow control gradient between on/off. Because velocity is not expanding. And there are a lot of artifacts. Mushrooms in explosions are the effect of the pressure (from what I understand), and as it's wrote in large Jeff post (in link above) : divergence field. But because I'm working right now on smoke solver only : it does not have this field (no combustion model). So main question is (probably): where should I control pressure. Or disable it? I've modified it on the dop level, with gasvop with pressure subtracted from itself, mult by 0 - but it did not had any effect (both presolve and postsolve). I assume it calculates this vel somewhere in between (?). I must be doing something wrong. 2. How to make smoke just keep going without twirls and all of that "nice stuff". Just inital vel, that's it. Imagine super simple sphere, default pyro, all off, only vel up. What I've tried: - somehow remove pressure (not sure I've done this right, but I could delete temperature, vel, density... but pressure without any luck) - advect vel field itself (no difference in vel = no pressure), but that lead to uncontrolled vel turbulence, when used with more complex emitter. - somehow blurring vel field and adding it when projectmultigrid was off to expand smoke. - make those red particles a volume and to look like smoke, but there is problem with disappearing (they kind of always pop out of existence even if particles fade nice and slow). Some tests: advecting smoke only with "pipevel" , again projectmultigrid off (no mushrooms = good)

It would be awsome to have some simple hip file. I don't know how to do that stuff . Yet.

Any luck Juraj? I'm trying to make ripping metal effect, but everywhere I look : "old technique, not working in H15". Are there any advanced FEM tutorials?

Well... it does slow down, but it orients bad. Anyway, friend found other solution: make blend and timeshift after transform, not before. Works well. I feel stupid a little.

I've already did that. Rop output or cache - does not work. sample file in attachment timeblend_and_transformPeaces.hipnc

Hallo! I've got big problem with jumping position and orientation. here it is (GREY: retimed animation, RED: transformed pieces) : https://youtu.be/lq_MwVrwUNU (sorry for the short video) Transform Pieces get's the job done when there is no retime, but when time blend comes in - it's just lost and does not take into account new values. Don't know how to force it to compute retimed animation into new frames. Like (making it up): @blendtime = @actualFrame; I'm not sure how to acces $F in vex. And bring it back. Or any other solution. timeblend_and_transformPeaces.hipnc

same here, same solution.

Aha... this is where my stupidity shines: add attrib in attribvop gives me probably average of all points, bind export gets me per point values. I don't understand why for now, but will get there someday. Probably it just creates attrib with some default value, not per-point attrib. Case closed. Slap in the ass earned.

Well... in conclusion: nevermind. I had to use foreach on points in previous system... had some problem with calculating length of all points, not individual ones. Was surprised then, but it only worked with foreach. Now I've checked pointvop and it's doing it's thing without foreach loop.