No clear winner. The survival gap is under 10 percentage points and the balanced-score gap is under 0.05, so neither SolvSRK nor the best baseline clears the win threshold. Either works — choose on cost, licensing, or integration effort. All verdicts →
15x15 woven fabric composite with checkerboard conductivity and gas pressure (dim=675, S3)
Kamal & Sourour (1973); Pojman (2012)
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.
def _kamal_sourour_rate_vec(T: np.ndarray, alpha: np.ndarray) -> np.ndarray:
T_safe = np.clip(T, _T_FLOOR, _T_CEIL)
alpha_safe = np.clip(alpha, 0.0, 1.0)
inv_RT = 1.0 / (_R_GAS * T_safe)
arg1 = np.clip(_E1 * inv_RT, 0.0, _EXP_ARG_MAX)
arg2 = np.clip(_E2 * inv_RT, 0.0, _EXP_ARG_MAX)
k1 = _A1 * np.exp(-arg1)
k2 = _A2 * np.exp(-arg2)
return (k1 + k2 * np.power(alpha_safe, _M)) * np.power(1.0 - alpha_safe, _N_ORD)
def _fp_woven_fabric_2d_rhs(t, y):
T_flat = np.clip(y[:_WF_N2D], _T_FLOOR, _T_CEIL)
alpha_flat = np.clip(y[_WF_N2D:2 * _WF_N2D], 0.0, 1.0)
P_flat = y[2 * _WF_N2D:]
T = T_flat.reshape(_WF_NX, _WF_NY)
alpha = alpha_flat.reshape(_WF_NX, _WF_NY)
dadt_2d = _kamal_sourour_rate_vec(T, alpha)
dT = np.empty((_WF_NX, _WF_NY))
for i in range(_WF_NX):
for j in range(_WF_NY):
kx = _WF_KX[i, j]
diff_x = _WF_DIFF_X[i, j]
diff_y = _WF_DIFF_Y[i, j]
ky = _WF_KY[i, j]
# x-direction Laplacian
if j == 0:
lap_x = 0.0 # Left edge Dirichlet
elif j == _WF_NY - 1:
T_ghost = T[i, j] + (_H_CONV * _WF_DX / kx) * (_T_AMBIENT - T[i, j])
lap_x = (T[i, j - 1] - 2.0 * T[i, j] + T_ghost) * _WF_INV_DX2
else:
T_left = _WF_T_LEFT if j == 1 else T[i, j - 1]
lap_x = (T_left - 2.0 * T[i, j] + T[i, j + 1]) * _WF_INV_DX2
# y-direction Laplacian
if i == 0:
T_ghost = T[i, j] + (_H_CONV * _WF_DY / ky) * (_T_AMBIENT - T[i, j])
lap_y = (T_ghost - 2.0 * T[i, j] + T[i + 1, j]) * _WF_INV_DY2
elif i == _WF_NX - 1:
T_ghost = T[i, j] + (_H_CONV * _WF_DY / ky) * (_T_AMBIENT - T[i, j])
lap_y = (T[i - 1, j] - 2.0 * T[i, j] + T_ghost) * _WF_INV_DY2
else:
lap_y = (T[i - 1, j] - 2.0 * T[i, j] + T[i + 1, j]) * _WF_INV_DY2
dT[i, j] = diff_x * lap_x + diff_y * lap_y + _SRC_COEFF * dadt_2d[i, j]
dT[:, 0] = 0.0
# Gas pressure
dP = (
(_RHO_RESIN * _V_GAS_SPECIFIC * dadt_2d.ravel() * _R_GAS_IDEAL * T_flat) / _V_PORE
- P_flat * _PERM_LOSS
)
dy = np.empty(_WF_DIM)
dy[:_WF_N2D] = dT.ravel()
dy[_WF_N2D:2 * _WF_N2D] = dadt_2d.ravel()
dy[2 * _WF_N2D:] = dP
return dyCanonical 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: high
Recommendation snapshot
Clean best: SciPy DOP853
Noisy best: SciPy BDF
Coverage
14 solver arms · clean + 5 noise levels
Ranked on survival, precision, and speed
20 seeds/cell default · 14 arms · TRL 4–5 · simulation-lab validated · this page: FP Woven Fabric 2D High-Res (fp-woven-fabric-2d-high-res)
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
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 DOP853SciPy | 100% | 13.2 | 12,362 | 5.59 s | 0.934 |
| 2 | SolvSRK | 100% | 10.6 | 30,845 | 11.65 s | 0.872 |
| 3 | SciPy RK45SciPy | 100% | 10.4 | 13,106 | 5.99 s | 0.866 |
| 4 | SciPy RadauSciPy | 100% | 10.3 | 19,356 | 11.42 s | 0.864 |
| 5 | Tsit5external | 100% | 10.2 | 13,842 | 8.68 s | 0.862 |
| 6 | SciPy RK23SciPy | 100% | 8.6 | 9,521 | 4.35 s | 0.823 |
| 7 | SciPy LSODASciPy | 100% | 7.7 | 25,719 | 11.71 s | 0.802 |
| 8 | CVODE Adamsexternal | 100% | 7.5 | 17,251 | 9.19 s | 0.797 |
| 9 | CVODE BDFexternal | 100% | 7.3 | 7,361 | 3.68 s | 0.793 |
| 10 | SciPy BDFSciPy | 100% | 7.2 | 4,176 | 2.56 s | 0.790 |
At Clean, best balanced arm is SciPy DOP853 · SolvSRK survival 100%, SCD 10.6.
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 →
SolvScout · free
This page shows one published benchmark cell. SolvScout fingerprints your ODE, compares it to the full corpus, and recommends a solver with the same survival / precision / speed ranking you see here — including when a SciPy arm wins.
SolvSRK · 30-day trial
SolvSRK is the stiffness-adaptive integrator behind the SolvSRK column in these tables. Create an account, activate a machine, and take a 30-day trial — same binary you'd ship after purchase.
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_fp_woven_fabric_2d_high_res_2026,
title = {Resonix Evidence Portal: FP Woven Fabric 2D High-Res},
author = {{Resonix Labs (Canada) Inc.}},
year = {2026},
howpublished = {\url{https://resonix.tech/evidence/problems/fp-woven-fabric-2d-high-res}},
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.}
}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