I made a test where I want to have wall been hit by an object (sphere in my case) and then the wall fractures at that spot and breaks apart, just to get familiar with RND stuff. So I set up my wall as glued fractured object, fractured as packed primitives. I set up a sphere as active rigid body, applied some initial velocity, so that it hits the wall.
So far so good, now comes the problem:
The glue value keeps the fractured pieces together, just like it's supposed to do, but when my sphere hits the wall it acts like a solid object and gets pushed away. So I tried to lower the glue value but then it's not strong enough to prevent the wall from collapsing on its own. When the sphere now hits the wall it breaks at the bottom and the top gets pushed away. Strangely parts that are already broken and detached from the rest of the wall a few frames later get draged with the rest of the object as if they were still one piece. This is very confusing. I can't describe it any clearer so I hope you understand what I mean when you see my hip file.
Keep collapsing wall in place
3323 1 0-
- anon_user_21411066
- Member
- 210 posts
- Joined: Jan. 2014
- Offline
-
- xukca
- Member
- 50 posts
- Joined: July 2013
- Offline
I checked your scene briefly:
-your wall is not clean enough. You must separate walls going on perpendicular directions from each other and then fracture them. Fracturing everything at once as you did leads to problems and funky geometry, not good enough for convex hull representation (check it in dops, and also default value of 0.2 for it is too big, try 0.05)
-your sphere is moving way too fast. This means the default substep of 1 is not enough. I highly advice to decrease the speed, bump the substeps of the whole sim a bit and increasing the density of the sphere, to achieve the effect you want.
-you probably need to either make the most lower parts of your wall inactive, or constraint them to the ground with hard constraint, so they stay anchored. This way you can increase the glue constraint strenght to somethhing better than 400 and your wall is not going to be siply dragged by the huge ball.
pesonal advice: learn to apply pin constraint networks. Simply using glue constraint networks is not good enough for realistic destruction. (fxphd, cgworkshop tutorials).
Keep going 8)
-your wall is not clean enough. You must separate walls going on perpendicular directions from each other and then fracture them. Fracturing everything at once as you did leads to problems and funky geometry, not good enough for convex hull representation (check it in dops, and also default value of 0.2 for it is too big, try 0.05)
-your sphere is moving way too fast. This means the default substep of 1 is not enough. I highly advice to decrease the speed, bump the substeps of the whole sim a bit and increasing the density of the sphere, to achieve the effect you want.
-you probably need to either make the most lower parts of your wall inactive, or constraint them to the ground with hard constraint, so they stay anchored. This way you can increase the glue constraint strenght to somethhing better than 400 and your wall is not going to be siply dragged by the huge ball.
pesonal advice: learn to apply pin constraint networks. Simply using glue constraint networks is not good enough for realistic destruction. (fxphd, cgworkshop tutorials).
Keep going 8)
-
- Quick Links