|
|
Problem with Phisics and Polygon modeling
About Truespace Archives
These pages are a copy of the official truespace forums prior to their removal somewhere around 2011.
They are retained here for archive purposes only.
Problem with Phisics and Polygon modeling // Roundtable
Post by ivampretti // Jan 30, 2009, 8:02pm
|
ivampretti
Total Posts: 151
|
Hi Folks, Im needing a little help here! I did an industrial thing called Convertedor, its like a bottle but is used to make steal. I modeled it starting with a shell cylinder and extrude comes extrude goes, tcharam we have a Convertedor. :D
I putted the "refratario"(yelow) inside the "Convertedor"(blue) and a ball(white). I want that the refratario fall down through the hole of the Convertedor. But looks like it has a face where should be a hole! I dont know why... take a look on the videos attached! I putted a ball just to test the hole!
Some body know what can I do with the Convertedor's mesh? |
Post by ivampretti // Jan 30, 2009, 8:17pm
|
ivampretti
Total Posts: 151
|
:banana::banana::banana: I found out! I am so stuped! the Convertedor was with the hole under the ground level!
But the things doesnt changed that much, the refratarios doesnt fall thru the hole, they stuck! I need to improve this! |
Post by early // Jan 31, 2009, 9:27am
|
early
Total Posts: 46
|
You can try making the pieces smaller, and play around with the elasticity and friction values.
What kind of converter are you modeling? Basic oxygen process converters, like the earlier Bessemer converters, have only a top opening, and the refractory lining was only replaced infrequently. (just curious.) |
Post by ivampretti // Jan 31, 2009, 7:58pm
|
ivampretti
Total Posts: 151
|
Yea thats the one,the basics type! But who designed this converter created a botton opening to change the refractory in an easy way. But the idea is not working, so now they are rebuilding the botton closing this opennig because they alredy change the refractary with the top hole and it works fine. Do you have any idea of how change the friction and elasticity parameters of all pieces at once?
Do you work with the steal factory? (just curious too :D ) |
Post by RAYMAN // Jan 31, 2009, 10:28pm
|
RAYMAN
Total Posts: 1496
|
I imagine you have problems with the physics with this one because the elements are stuck....
I deleted every second element and added a local wind.
Just didnt tune up the resistence enough though and have some elements
passing through the converter but nevertheless with local enviroment and wind I got the elements out.
Peter |
Post by RAYMAN // Jan 31, 2009, 11:54pm
|
RAYMAN
Total Posts: 1496
|
I have always opted in for a stack for Ts like Vue has where you can group
your objects and apply shaders or physical properties to a whole group....
I think both layer systems dont support it but I am not sure but i doubt it !
Peter |
Post by early // Feb 1, 2009, 10:28am
|
early
Total Posts: 46
|
I like the idea of using wind to force the pieces out, but it would all depend if the goal is to be realistic or not. But if you're just demonstrating the process and not looking for a 'physical simulation', it could be effective. Or maybe use a local environment stream.
Group the pieces together and then assign the material type. Make the grouped object breakable. Then you can break the object and let the pieces either fall by gravity, get blown out by the wind, or hosed out by a stream.
The only thing I really know about steelmaking is just general knowledge type stuff. Outside of computers, I wouldn't consider myself an expert at anything, but I'm curious about everything. |
Post by ivampretti // Feb 1, 2009, 11:42am
|
ivampretti
Total Posts: 151
|
I liked the idea of local wind, but Peter, I need the pieces come out through the opening on the top! Because the refractory will be broken and then the converter will rotate and the trash will need to be cleaned! Im doing this because I sold an animation about the changes on the bottom! The refractory caming out through the top opening is my desire :D just to see it more beaultifull on the screen! I need this solved until february 4th. So I have sometime to make this better :p
I'll try to make smaller pieces and find a way to change the physical parameters of all at once, I dont know how yet. I'll post the result here in a couple hours. |
Post by ivampretti // Feb 1, 2009, 4:00pm
|
ivampretti
Total Posts: 151
|
I discovered two cool things! First if you have an encapsulated number x of objects, you can set the physical atributtes for each object at once. And there is an option on the PhysObject panel called BreakAble. This allow a group to break in his separated pieces. Awesome :jumpy:
The other thing is that is too much information to TS process the way that I want. It takes too long to make the things happend because it stay calculating the colisions between the internal group's objects.
So I was wondering if there is an way of turn of the colision between the objects of the same group and just collide with the Converter. The refractory can move through each other, I just want that they have colisions with the Converter. then I bealive its possible to use as many pieces as I want. In the PhysObject panel there is an SelfColision enable, but seems only work with characters skeleton.
AnyBody? |
Post by TomG // Feb 2, 2009, 4:37am
|
TomG
Total Posts: 3397
|
I'm unaware of a way of doing that, if something has physical properties it will interact with anything else with physical properties.
Why not run the sim and record the animation for one piece at a time? That way each piece will only interact with the surrounding container, and then its animation will be stored in the KFE, and you repeat for the next piece.
Whether this proves faster than letting them collide with each other, I don't know. It would make it easier to adjust the results of one piece until you were happy.
Of course they will them "fall through each other" as they come out, since they aren't interacting, so that could be a problem. You could stagger each KFE animation for them to emerge at different times and not intersect perhaps.
HTH!
Tom |
|
