Dude, that is soooooooooooooo meta.
Simulations of physics are simple in the sense that all you're really doing is evaluating equations.
Let's just consider classical physics for the time being.
To simulate a collection of atoms all you really have to do is calculate the electric interactions between them and apply that to their motions as that's what creates forces between them. So, you'd have the classical view of the atom which is the circular electron clouds with hard sphere electrons. There's voltage and field equations so you can apply that calculating forces and then velocities, etc.
If you want, you can even include chemistry and have atoms form molecules when certain conditions are met.
Considering we have a good chunk of science completed, it is entirely possible to simulate the equatons, using the proper structures.
Like, an electron for example would be like this :
Code:
struct electron {
unsigned mass;
float x, y, z; // position
float vx, vy, vz; // velocity
float ax, ay, az; // acceleration
float charge;
float kinetic_energy, potential_energy;
};
and I think that encompasses most of a classical electron.
Using structures to represent particles should allow you simulate physics. The real problem is just that it's not feasible considering current technology.
Edit : Plus, aren't you like an EE/CE kind of guy? You should know how computers are actually built. Just write code that would simulate that. So it'd be current flows and all that good stuff.