It definitely increases. I’m running the test after the fix.
Likely Culprit:
With PGS we use a “collision resolution” trick when two objects are caught interpenetrating. Interpenetration is a natural phenomenon in discreet physics engines, so the new solver basically takes the part and moves it out of the collision (without adding any velocity). In a system like this, where majority of the collisions are with the floor (due to gravity and so on, and the starting state), by displacing the part upwards, you are actually adding potential energy into the system.
So it’s possible/most likely that the energy slowly increases due to this collision resolution step. It introduces extra potential energy.
Eliminating the balls that shoot out of the container might be a good idea too, might cause weird drops like that.
EDIT: And do what someone suggested, throw in bodyforces to turn gravity off, and initialize the system with more kinetic energy on one side (instead of more potential energy). Just to see if the penetration solving is the major issue here.
[ul]
[li]Energy leaks were caused when balls would leave the system. This was a bug with Ball-Block collisions, and we have a fix for it that will be out by mid Monday.[/li]
[li]Energy increase was caused by positional correction + gravity. Positional correction can alter energy state in any potential fields, and gravity is a constant potential field.[/li]
[/ul]
I made all the parts have 0 gravity and found that the energy only increases at 0.05% over a course of 15 minutes. Here is the updated code (Please note, this will probably leak energy until we make the Ball-Block fix public):
local AllBalls = workspace.AllBalls
local LG = BrickColor.new("Lime green")
local BG = BrickColor.new("Bright green")
local C = BrickColor.new("Camo")
local O = BrickColor.new("Olive")
local NO = BrickColor.new("Neon orange")
local MR = BrickColor.new("Medium red")
local BR = BrickColor.new("Bright red")
local totalEnergy = 0
wait(7)
function getTotalEnergy(part)
local KE = ((part.Velocity.magnitude^2)*(part.Size.X^3)/70000)/2
--local PE = ((part.Position.Y - 5) * 196.2) *(part.Size.X^3)/70000
return KE-- + PE
end
for i,v in pairs(AllBalls:GetChildren())do
totalEnergy = totalEnergy + getTotalEnergy(v)
end
for i,v in pairs(AllBalls:GetChildren())do
bf = Instance.new("BodyForce", v)
bf.force = Vector3.new(0, 196.2, 0) * v:GetMass()
v.Anchored = false
end
print("ENERGY SHOULD INCREASE")
--game.Workspace.AllBalls.Driver1.Velocity = Vector3.new(200, -1, 200)
--game.Workspace.AllBalls.Driver2.Velocity = Vector3.new(-100, -1, -100)
print("Average Energy: " .. tostring(totalEnergy/#AllBalls:GetChildren()))
counter=1
while wait(0)do
for i,v in pairs(AllBalls:GetChildren())do
if i%200==0 then wait(0)end
local s = (v.Velocity.magnitude^2)*(v.Size.X^3)/70000
if s<8 then
v.BrickColor = LG
elseif s<16 then
v.BrickColor = BG
elseif s<25 then
v.BrickColor = C
elseif s<45 then
v.BrickColor = C
elseif s<85 then
v.BrickColor = O
elseif s<100 then
v.BrickColor = NO
elseif s<130 then
v.BrickColor = MR
elseif s>180 then
v.BrickColor = BR
end
counter = counter + 1
totalEnergy = 0
if counter%200==0 then
counter = 0
for i,v in pairs(AllBalls:GetChildren())do
totalEnergy = totalEnergy + getTotalEnergy(v)
end
print("Average Energy: " .. tostring(totalEnergy/#AllBalls:GetChildren()))
end
end
end
I told you we haven’t fixed the bug where balls magically leave the system. This will be available to you on Monday. You’re seeing energy drops because balls magically escape the container, and then they get deleted, and no longer are part of the energy of the system.
[quote] Am I a fool for not understanding any of this :Q.
I’m taking physics this year so maybe I’ll be able to understand it here in a few weeks lol [/quote]
You’d be better off and learn faster with videos on the internet or coursera.com
lol