Ok, I have a problem I''m trying to figure out. Say you start off in an environment with no gravity whatsoever. Then, you have atoms of different masses floating around. What I know is that everything exerts a gravity field albeit a very small in most instances. So atoms should exert a small gravity field upon each other and eventually attract. Knowing this, how would I impliment this using a language like C++ or say Matlab? Is there some kind of ratio that determines the amount of gravity a body exerts? Also, is there a way to implement it with an environment with existing gravity? Thanks for the help.

F_g = (Gm_1m_2)/r^2

Here, G being the universal gravitational constant (roughly 6.67e-11 N*m^2/kg^2), m_1 and m_2 being two masses and r being the distance between them. Since gravity is a force of attraction, it''s mostly meaningless without at least two point masses.

However, you may be referring to the gravitational field (a concept very similar to those of the electrical and magnetic fields, and was in fact taken from these), in which case:

g = F_g/m

Where in this instance, g is a vector (NOT the acceleration due to gravity) and m is a "test mass". So you could also write it as such:

g = (GM/r^2)*rhat

M being the mass of the object you''re interested in and rhat being the unit vector of r to give direction (since I don''t know how to actually draw one of those in HTML ).

I suggest you pick up any intro physics textbook and read the chapter on gravity (there''s almost always a complete chapter devoted to it). It should clear up any other issues you may have.

I was reading an article on Quantum Physics. What i understood is that these Newtonion Physics isnt accurate at atomic level.

This page explains the four fundamentsl forces in physics, and why gravity doesn''t matter as much on small scale systems.

http://hyperphysics.phy-astr.gsu.edu/hbase/forces/funfor.html

quote:

---

Original post by DirectXXX
I was reading an article on Quantum Physics. What i understood is that these Newtonion Physics isnt accurate at atomic level.

---

What does''t hold is mechanics, gravity still holds at atomic level. However, seen the very little distances and the huge strong nuclear and EM forces, gravity is trascurable

Saying mechanics doesn''t hold isn''t entirely accurate. The truth is, Newtonian mechanics is in fact an approximation to begin with on a macroscopic scale. It just happens to be a very close approximation when dealing with everyday objects. On extremely large scales and extremely small scales, the approximations are far less accurate.

And yes, since gravity is one of the four fundamental forces of nature, it most definately applies on the atomic level. But since the value of G is so small (not to mention the masses of atoms and even smaller subatomic particles), it approaches zero and can, for most purposes, be completely ignored since it is dwarfed by the other three forces (at that level).

So if I''m creating a program simulating these 4 fundamental forces and their effects on other atoms and molecules I would want to include all of them for an accurate model, but could sacrifice gravity because of its very small effects right?

Also, how hard would if be to simulate this? Or has it already been done?

At single- or sub-atomic scales, the model of atoms and their parts (protons, neutrons and electrons) as discrete points is not accurate. At that scale, you need to start using quantum mechanics to calculate the propegation of a system. This requires rather advanced university level courses for anything more than a single electron hydrogen atom.

At slightly larger scales, which is probly what you''d like to simulate, you can probably treat atoms as dipoles (polar molecules), non-interacting particles (nonpolar molecules, neutral atoms) or charged particles (ions, free radicals). For dipole interactions to work, you''ll need to keep track of the angular orientation of the molecules and determine how the dipoles rotate (torque) due to other particles'' dipole forces as well as the translational effects. Direct Particle collisions can be treated as hard elastic collisions, I think... though atoms do tend to bond together if they possess sufficient activation energy...

quote:

---

Direct Particle collisions can be treated as hard elastic collisions, I think... though atoms do tend to bond together if they possess sufficient activation energy...

---

You are correct. So actually, you could have either completely elastic or completely inelastic collisions (ignoring the loss of energy for the in between ones). But for the most part, you could model it such that a majority of the collisions were elastic.

so per-pixel collision detection wasn''t enough for you, you''re going with per-atom? My game is way better, I have per-neutron hit ditection and per-neutrino lighting.