The Drude model treats conduction electrons as a classical gas that flies freely between random collisions with the vibrating lattice ions. Between collisions each electron accelerates under the field E; at a collision its velocity is randomised, erasing the memory of that acceleration. Averaged over the whole gas this gives a small net drift on top of a much larger, essentially random, thermal motion:
a = eE/m (free acceleration between collisions)
v_d = eEτ/m (mean drift velocity)
J = n·e·v_d = σE (current density, Ohm's law)
σ = n·e²·τ/m (conductivity)
τ(T) ≈ τ₃₀₀·(300/T) (scattering time falls as the lattice gets hotter)
- n — conduction-electron density (fixed per metal, from its known valence and atomic density).
- τ — mean free time between scattering events, back-derived per metal from its measured room-temperature resistivity, then scaled with temperature.
- Metal — swaps n and τ₃₀₀: silver scatters least (longest τ, highest σ), iron scatters most.
- E field / Temperature sliders and the Reverse Field button feed directly into the equations above — the drift velocity, current density and conductivity readouts are computed analytically from them, not measured off the animation.
- The 3D scene shows the same mechanism qualitatively: electrons (small spheres) fly ballistically, scatter off the fixed ion lattice at a rate set by 1/τ, and pick up a slight bias toward the field direction between scatters. In reality that drift is thousands of times smaller than the electrons' thermal speed — the animation exaggerates it so the bias is visible.