Particle Flow Discussion

#5655

Cool, that came out nice! now I see why you had soo many materials :slight_smile:

#5656

Lookin good Ansi

#5657

Cool effect Ansi :slight_smile:

Some box #2 R&D on metal denting on a cube.

And a working drawbridge that lowers with chains (box #2) - did a trick here to avoid using a chain link in the sim, and instead used a cylinder shape, then changed the Shape with a new Shape: a mesh of two chains.

@ Glacierise - it should work with any Particle Skinned mesh, but having it shake a bit is nice, so it nudges. Works ok for static objects, but now trying to find a way to nudge the cube

#5658

Seriously awesome stuff Robert! How stable and useful would it be on a heavy animated mesh?

#5659

@Robert: That’s sweet! Especially the ones that get trapped :smiley: Do you glue the spheres, break by force and re-glue in a next event? You just gave me a nice idea about time lapse weathering effects with your point cached denting without the balls impacting :smiley:

@ Glacierise: I would always do the simulation part on a low res mesh that is a tid bid bigger then the high res version (i usually use a push modifier and inflate it a bit) and skin the high res mesh to it once the particles are cached with “Inside Inclusion” checked, that way 3 low res particles can influence a whole area of multiple high res objects with a falloff radius. PFlow never “sees” the actual highres mesh during calculation that way. Or use low res fragments or what ever you are doing, get the motion etc. done and switch em out after caching within cache disk’s Post Cache Operators. So far no one could tell a sim was done with less resolution meshes. In my ripping test i did once the statue had 1.7 million polys and the most enduring part was actually waiting on the particle skinner to be done skinning.

#5660

Ok, but how will you transfer the topology changes, the tearing overall, from the low to the hi-res mesh?

#5661

That looks very cool. Nice research.

#5662

thanks guys :smiley:

@ Ansi - yea man, in fact your cloth rip gave me the idea, but instead pipe the glued wire of broken bits off the skin into a new event, where there is a Switch that locks the particles in place, but this caused the “cube” to not move either,

#5663

nice, would love to check it out :slight_smile:

#5664

Hehe cool work Rob, thinking outside the Box :smiley:

#5665

thanks - here’s the max 2010 scene :twisted:

would it be hard to have box #3 detect a particle that breaks off, and assign it only gravity?

#5666

dont think it would be hard, let me test that. Glue can report all data to a data operator and writes it into a data channel that you can then butcher in a data op. But i’m pretty sure you wouldnt even need box3 for this, maybe try a physx switch set to anti-gravity in the first event and then you can you test “break all binding” true and pass it into a new event with gravity?

EDIT: thats how you did it pretty much :smiley:
EDIT: I mocked something up super quick, maybe it’s stooopid to do it that way, maybe you can pick at it and it is useful. I went the route and detect when a particle collided with another particle (in your case the ball). In the physx shape i set the Mass to the most minimum value and applied anti gravity via a physx switch. The very important thing is to check “Report Data to Operator” in the intercollision test :smiley: That took me a good 5 minutes to figure out LOL. Otherwise the physx Input set to detect intercollision wouldn’t do anything ;-p Anyways i use that as an ON/OFF switch to apply a new mass based on the particles size, color it differently and apply some gravity. The setup needs some love on the gravity side but at least it works in one event, no test necessary(other then the intercollision).

#5667

sweet - messing with your PFlow, and the Box #3 ops are helpful in better Box #2 control :wise:

#5668

hi!
playing with box2… trying to recreate the metal deformation/plasticity effect shown in the 5th minute 59th second of this video http://www.orbaz.com/products/particleflo/box2/Box2ProDemoVideo.html
here is my current result and setup (its far far away from the expected result :frowning: )…

the setup consists of 4 events the “inner squares” event contains a grid of cuboids glued together to form a larger square metal piece which is held in place by the “border squares” event which contains a number of cuboids arranged to form a frame around the “metal piece” these cuboids are glued only to the “inner squares” and do not participate in the physX simulation (therefore the event contains physX switch operator named “Freeze” to turn of the simulation)

however i have 2 problems with this setup :

  1. at the moment i have timed the setup so i know that at frame 5 the bindings between the “border squares” and the “inner squares” break (due to collision with the Ball event) therefore i used an age test to send the “inner squares” to the event dent motion where i have increased the max binds per particle of the glue operator so that the bonds between the inner squares become stiff so that there shape is permanently deformed … i would like to replace this with a system which forwards the particles from the inner squares event automatically to the dent motion event when the bindings between the border squares and the inner squares break this way i do not have to re-time or re calculate the value for age test whenever i change the speed or density of the ball

2)how to tone down the berserk spinning of the metal piece towards the end of the simulation?

i tried using group selection and split group test to solve the first problem but failed :frowning: …

[attached max 2010 file]

thanks! :slight_smile:

#5669

@ HornBerger - nice file :thumbsup:

Sand & Debris Blasts
Here’s a max 2010 edit, from a cool sand & dust file Allan made and shaded.

You will find it handy for sand, lava, sparks, shaprnel, fire-fights and splashes.
i changed flow to let us shot dynamic, to do multi-blast effects – great with Krakatoa. :twisted:

#5670

without checking the file it looks like the peak of spawn thing Andrew made a while ago :slight_smile:
http://www.orbaz.com/forum/viewtopic.php?t=1427

super helpful!

#5671

hey!
i couldn’t figure out an “out of the box” solution for the first problem i mentioned in the previous post

i would like to replace this with a system which forwards the particles from the inner squares event automatically to the dent motion event when the bindings between the border squares and the inner squares break this way i do not have to re-time or re calculate the value for age test whenever i change the speed or density of the ball

so i decided to use script operators with a persistant global variable VAL (its not the most elegant solution… but it is one :slight_smile: ) here’s the screen grab of the setup :

still haven’t figured out a way to reduce the chaotic spin motion the metal piece gets into once the balls rolled over it though :frowning:

Sand & Debris Blasts
Here’s a max 2010 edit, from a cool sand & dust file Allan made and shaded.
You will find it handy for sand, lava, sparks, shaprnel, fire-fights and splashes.

but the file only contains the sand Particle flow system (cool effect with a simple setup though! :thumbsup:) is the rest of the file available for download or is it a part of some video training/class or something

thanks! :slight_smile:

#5672

The question how to get 3D motion blur with changing topology especially in Glu3D, PWrapper and Particles comes up a lot in forums so here is a quick round up on how to get it to work. It’s a simple setting that’s sometimes overlooked… :wink:

Vimeo Link: 20:44minutes, 1680x916, 278.09MB
http://vimeo.com/14900721

cheers!
Ansi

#5673

Cool Anselm, good stuff :slight_smile:

#5674

still haven’t figured out a way to reduce the chaotic spin motion the metal piece gets into once the balls rolled over it though :frowning:

@Hornberger, use a PhysX Drag and hammer it, use a value like 10.0 and 10.0 on event entry, play with the speed multiplier too.