A reusable tile TPS (Space Shuttle-style LI-900 silica) works the opposite way to an ablative shield: it does not erode. Instead its very low thermal conductivity keeps almost all of the incoming heat near the outer surface, where a thin high-emissivity coating radiates it back to the sky. What does get through keeps conducting inward for a long time after the aeroheating stops — the "soak-back" that this simulator reproduces.
The tile is modelled as a 1D slab, solved with explicit finite differences:
∂T/∂t = α · ∂²T/∂x², α = k / (ρ·c_p)
Surface (radiative equilibrium):
q"_incident(t) − ε·σ·(T_s⁴ − T_∞⁴) = −k·(∂T/∂x)|_surface
Backface (coupled to airframe/cabin):
k·(∂T/∂x)|_back = h·(T_back − T_cabin)
q"_incident(t) follows a reentry-like pulse, rising and falling as a half sine over ~150 s at the peak flux you set. σ = 5.67×10⁻⁸ W/m²K⁴ is the Stefan–Boltzmann constant; ε is the coating emissivity you control — a lower ε means less radiative cooling, so the surface runs hotter for the same flux.
- Peak heat flux — the aeroheating rate at the trajectory's hottest point (Shuttle-class entries peaked near 25–35 W/cm²).
- Insulation thickness / conductivity — thicker or lower-k insulation slows conduction, so less heat reaches the structure before the pulse ends, but the stored heat still has to go somewhere: it keeps creeping inward afterward.
- Thermal penetration depth ≈ √(α·t) — roughly how far the heat wave has diffused into the tile; when it is much smaller than the tile thickness, the structure barely notices the pulse in real time.
- Watch the backface temperature keep climbing after the surface has already cooled — the peak structural temperature on a real orbiter tile can occur many minutes after touchdown, which is exactly why post-landing "cool-down" time matters for turnaround.