This is a from-scratch 2D reconstruction, not a flattened 3D view. The top panel is a real kinetic-theory gas simulation: hundreds of colliding CH₄ and H₂ disks obey elastic 2-body collisions (momentum and kinetic energy exactly conserved every collision). Because equipartition gives v ∝ 1/√m at equal thermal energy and H₂'s molar mass is ~8× lighter than CH₄'s, the cyan H₂ disks visibly dart ~2.8× faster than the orange CH₄ disks — a real consequence of kinetic theory, not a decoration.
v_H2 / v_CH4 = √(M_CH4 / M_H2) = √(16/2) ≈ 2.828
LHV_blend = x·LHV_H2 + (1−x)·LHV_CH4
ρ_blend = x·ρ_H2 + (1−x)·ρ_CH4
Wobbe Index = LHV_blend / √(ρ_blend / ρ_air)
The bottom-left panel solves Fick's second law ∂C/∂t = D·∂²C/∂x² by explicit finite differences across the pipe-wall thickness, with a Sievert's-law surface boundary C(0,t) = k·√(P·x) (atomic hydrogen solubility scales with the square root of its partial pressure). Integrating the resulting concentration profile gives a wall H₂ inventory that grows over time as exposure continues — a genuine time-dependent diffusion result, not an instantaneous multiplication. Composite/polymer pipe uses a far lower diffusivity and solubility constant, so its profile barely penetrates the wall.
- H₂ blend fraction — sets the probability a re-injected gas particle is H₂, and sets the diffusion boundary concentration via Sievert's law.
- Pipeline pressure — raises the diffusion boundary concentration (∝ √P) and nudges the chamber's drift speed.
- Flow speed — sets both the chamber drift velocity and the wall-exposure time compression.
- Pipe material — switches the diffusivity/solubility constants used by the finite-difference solver.