Hypersonic Aero-Thermal Coupled (dim=13)
ADVANTAGES2 · dim 13SolvSRK wins. At the comparison noise level, SolvSRK beats the best baseline by at least 10 percentage points of survival, or by at least 0.05 balanced score when survival is tied. Use SolvSRK for this class of problem. All verdicts →
Coupled 3-DOF trajectory and 4-layer TPS thermal response. State: [x,y,z, vx,vy,vz, T_surface,T_mid,T_bondline,T_substrate, q_stag_norm, alpha_deg, mass]. Sutton-Graves stagnation-point heating coupled to lumped TPS conduction with radiative cooling (epsilon=0.85). Stiff from disparate thermal/mechanical timescales.
Problem definition
Sutton & Graves, NASA TR R-376 (1971); Tauber & Sutton, J. Spacecraft 28(1) (1991); MIL-HDBK-5 TPS material properties
Canonical RHS excerpt from the registered callable used for this benchmark cell. Expand it to verify the state equations; it is not a standalone runnable fixture.
Show canonical RHS excerpt
def _atmosphere_density(altitude):
"""Exponential atmosphere model."""
h = max(altitude, 0.0)
return _RHO0 * np.exp(-h / _H_SCALE)
def _gravity_accel(pos):
"""Spherical gravity: returns acceleration vector (toward center)."""
r = np.linalg.norm(pos)
r = max(r, _R_EARTH * 0.5)
g_mag = _MU_EARTH / (r * r)
return -g_mag * pos / r
def _mach_number(speed, altitude):
"""Approximate Mach from speed; sound speed decreases with altitude."""
# Simplified: assume isothermal atmosphere for sound speed
h = max(altitude, 0.0)
a = _SPEED_OF_SOUND_SL * np.exp(-h / (2.0 * 42000.0))
return speed / max(a, 1.0)
def hypersonic_aero_thermal_coupled_rhs(t, y):
pos = y[0:3]
vel = y[3:6]
T = y[6:10].copy()
T = np.maximum(T, 200.0)
mass = max(y[12], 100.0)
speed = np.linalg.norm(vel)
speed = max(speed, 1e-6)
v_hat = vel / speed
altitude = np.linalg.norm(pos) - _R_EARTH
rho_inf = _atmosphere_density(altitude)
mach = _mach_number(speed, altitude)
cd = np.interp(mach, _MACH_TABLE, _CD_TABLE)
q_dyn = 0.5 * rho_inf * speed * speed
drag = q_dyn * cd * _S_REF
f_drag = -drag * v_hat
g_accel = _gravity_accel(pos)
if t < _T_BURN and mass > _M0 - _MDOT * _T_BURN:
f_thrust = _THRUST * v_hat
dm_dt = -_MDOT
else:
f_thrust = np.zeros(3)
dm_dt = 0.0
accel = (f_drag + f_thrust) / mass + g_accel
# Sutton-Graves stagnation-point heating
rho_inf_safe = max(rho_inf, 1e-20)
q_stag = _SUTTON_GRAVES_COEFF * np.sqrt(rho_inf_safe / _R_NOSE) * speed**3
# TPS thermal response: 4-layer lumped conduction
# Surface BC: net heat flux = stagnation heating - radiative cooling
q_rad = _TPS_EPSILON * _SIGMA_SB * T[0]**4
q_net_surface = q_stag - q_rad
# Inter-layer conduction fluxes (positive = toward substrate)
q_01 = _TPS_KEFF_01 * (T[0] - T[1]) / _TPS_DX_01
q_12 = _TPS_KEFF_12 * (T[1] - T[2]) / _TPS_DX_12
q_23 = _TPS_KEFF_23 * (T[2] - T[3]) / _TPS_DX_23
dTdt = np.empty(4)
dTdt[0] = (q_net_surface - q_01) / _TPS_RHOC[0]
dTdt[1] = (q_01 - q_12) / _TPS_RHOC[1]
dTdt[2] = (q_12 - q_23) / _TPS_RHOC[2]
dTdt[3] = q_23 / _TPS_RHOC[3] # insulated back face
# q_stag_norm tracks normalized stagnation heat flux (for diagnostics)
q_peak_ref = 5e6
dq_norm_dt = (q_stag / max(q_peak_ref, 1.0) - y[10]) / 10.0 # relaxation
dy = np.empty(13)
dy[0:3] = vel
dy[3:6] = accel
dy[6:10] = dTdt
dy[10] = dq_norm_dt
dy[11] = 0.0 # alpha_deg held constant
dy[12] = dm_dt
return dy- Parameters
- _CD_TABLE = [0.3, 0.35, 0.25, 0.15, 0.12, 0.11, 0.1, 0.1]
- _H_SCALE = 8500
- _M0 = 1500
- _MACH_TABLE = [0, 1, 2, 5, 10, 15, 20, 25]
- _MDOT = 50
- _MU_EARTH = 3.986e+14
- _RHO0 = 1.225
- _R_EARTH = 6.371e+06
- _R_NOSE = 0.1
- _SIGMA_SB = 5.67037e-08
- _SPEED_OF_SOUND_SL = 340.29
- _SUTTON_GRAVES_COEFF = 0.00017415
- _S_REF = 0.5
- _THRUST = 200000
- _TPS_DX_01 = 0.0075
- _TPS_DX_12 = 0.0065
- _TPS_DX_23 = 0.0115
- _TPS_EPSILON = 0.85
- _TPS_KEFF_01 = 0.190476190476
- _TPS_KEFF_12 = 0.166666666667
- _TPS_KEFF_23 = 0.975609756098
- _TPS_RHOC = [9000, 1600, 4320, 27000]
- _T_BURN = 30
- Initial condition
- y(0) = [0, 0, 6.401e+06, 3000, 0, 300, …] [shape=(13,), min=0, max=6.401e+06]
- Horizon
- t ∈ [0, 600]
Canonical RHS excerpt captured from the same registered callable used for the published benchmark. Frozen closure values are summarized below; helper imports and solver settings are intentionally omitted.
Fingerprint
Spread: extreme
Default noise: medium
Recommendation snapshot
Clean best: SciPy BDF
Noisy best: SolvSRK
Coverage
14 solver arms · clean + 5 noise levels
Ranked on survival, precision, and speed
Versions & freeze
Methodology →- Freeze
- 2026-08-13
- libsolvsrk
- 2.3.0
- SciPy
- 1.14
- SUNDIALS
- CVODE (bundled backend)
20 seeds/cell default · 14 arms · TRL 4–5 · simulation-lab validated · this page: Hypersonic Aero-Thermal Coupled (dim=13) (hypersonic-aero-thermal-coupled-dim-13)
Governed SolvTune benchmark freeze; per-arm medians only. RHS definitions and raw trial rows are not published.
Self-reported by Resonix Labs · not independently verified
Results matrix
Pick an objective and a noise level to rank all arms on survival, median SCD, median nfev, and median wall time. Medians across seeds.
Objective
Best overall trade-off of survival, precision, and speed.
Noise level
| # | Solver | Survival | SCD | nfev | Wall | Score |
|---|---|---|---|---|---|---|
| 1 | SciPy BDFSciPy | 100% | - | 1,860 | 53 ms | 0.809 |
| 2 | SciPy RadauSciPy | 100% | - | 4,460 | 97 ms | 0.809 |
| 3 | SciPy RK45SciPy | 100% | - | 2,300 | 31 ms | 0.809 |
| 4 | SciPy LSODASciPy | 100% | - | 1,261 | 19 ms | 0.809 |
| 5 | SciPy DOP853SciPy | 100% | - | 2,246 | 30 ms | 0.809 |
| 6 | SciPy RK23SciPy | 100% | - | 4,385 | 95 ms | 0.809 |
| 7 | CVODE BDFexternal | 100% | - | 1,097 | 35 ms | 0.809 |
| 8 | CVODE Adamsexternal | 100% | - | 1,021 | 35 ms | 0.809 |
| 9 | Tsit5external | 100% | - | 2,208 | 996 ms | 0.809 |
| 10 | SolvSRK | 100% | - | 3,330 | 90 ms | 0.809 |
At Clean, best balanced arm is SciPy BDF · SolvSRK survival 100%.
Values are medians across seeds, measured by Resonix Labs on Resonix hardware and not independently verified; nfev and wall are on reference lab hardware (indicative). Under injected noise only SolvSRK and the SciPy arms are run. How we measure accuracy → · Verification status →
Cite this page
Replace the access date. Pin the freeze ID and library versions when comparing against a later export. Cite it as what it is - a self-reported vendor benchmark, not an independently verified result. The note field says so; please keep it.
@misc{resonix_evidence_hypersonic_aero_thermal_coupled_dim_13_2026,
title = {Resonix Evidence Portal: Hypersonic Aero-Thermal Coupled (dim=13)},
author = {{Resonix Labs (Canada) Inc.}},
year = {2026},
howpublished = {\url{https://resonixusa.com/evidence/problems/hypersonic-aero-thermal-coupled-dim-13}},
note = {Self-reported vendor benchmark; internally generated by Resonix Labs and not independently verified. Accessed YYYY-MM-DD. Freeze 2026-08-13; libsolvsrk 2.3.0; SciPy 1.14.}
}Related
TRL 4–5 · simulation-lab validated · 398 problems · 14 solver arms · clean + 5 noise levels
Freeze: 2026-08-13 · scipy 1.14 · libsolvsrk 2.3.0 · Methodology
Self-reported by Resonix Labs · not independently verified · Verification status