Verification

How we verify STRIX

STRIX is an independent solver built to be correct, then confirmed against known references. This page shows the methodology and representative cross-verification results — the engineering facts are open; the presentation is our own.

21/21 international benchmarks passed KDS · KBC · KCI · FEMA OpenSees (UC Berkeley)

Download full V&V report (PDF) Cross-code verification note (PDF)

Methodology

Four ways we keep results honest

🧮

OpenSees foundation

Built on the UC Berkeley OpenSees finite-element solver — a world-standard engine — rather than an opaque in-house black box.

📐

Converged to theory

Results are mesh-converged against analytical and code-theoretical solutions — STRIX aims to be correct, not tuned to match any one competitor.

🔗

Standards-traceable

Every design value traces back to its governing combination and KDS/KCI/FEMA clause through Result DNA.

⚖️

Independent cross-checks

Modal periods, member forces and design ratios are cross-verified against established reference models (MIDAS / ETABS).

Verification

We earn trust with data,
not logos.

21 international benchmarks, four categories

Every case below runs the real production code path (builder → OpenSees engine → parser), headless, against a published reference solution — not an internally-invented target. Element, analysis and nonlinear behavior are each checked against a different independent source.

NAFEMS CSI SAP2000 Verification Manual ASME Timoshenko & Roark closed-form

Element Benchmarks

11 benchmarks

Benchmark	STRIX	Reference	Δ
SB1 — Tip deflection u_z · Euler-Bernoulli 1D cantilever	−0.107865 mm	−0.107865 mm	+0.00020%
SB2 — Tangential edge stress sigma_yy at D · NAFEMS LE1 elliptic membrane	90.8514 MPa	92.7 MPa	−1.994%
SB3 — Normalized tip displacement u_QEt/P (mid loaded edge) · Cook's membrane	23.9578 norm.	23.91 norm.	+0.2%
SB5 — Deflection coefficient α · Rectangular plate bending — thin plate, uniform & central point load	0.00406	0.00406	+0.064%
SB6 — Deflection coefficient α · Thick rectangular plate bending — transverse shear, uniform load	0.004271	0.004273	−0.057%
SB7 — Center deflection Uz · Simply-supported beam on a Winkler elastic foundation	−0.089333 in	−0.089333 in	+0.000057%
SB8 — Natural frequency f₁ · Deep simply-supported beam — shear-flexible (Timoshenko) natural frequencies	102.149374 Hz	102.149414 Hz	−0.000039%
SB9 — Midspan deflection Uz — combined (axial+bending) · Rigid steel portal frame under UDL — bending + axial deformation	−69.3655 mm	−69.372833 mm	+0.0106%
SB10 — Brace E1 axial N · Asymmetric determinate two-bar truss	10000 N	10000 N	0.00%
SB12 — Oblique Ux · Elastic Link (twoNodeLink) Beta Angle coordinate transformation — inclined 6-DOF spring	0.152961 mm	0.152961 mm	+0.000081%
PD1 — Uz (w/ tension) · Tension stiffening via P-Delta analysis	−0.543496 in	−0.543305 in	−0.0352%

Analysis Benchmarks

6 benchmarks

Benchmark	STRIX	Reference	Δ
SM5 — Eigenvalue ω² · Bathe & Wilson eigenvalue problem — ten-bay nine-storey plane frame	0.589538 rad²/s²	0.589541 rad²/s²	−0.00050%
SM5b — Eigenvalue ω² · Rigid-diaphragm eigenvalue condensation — eccentric multi-storey building	2070.118745 rad²/s²	2070.117026 rad²/s²	+0.00010%
SM6 — Eigenvalue ω² · ASME eigenvalue frame — 3-D fixed-base pipe frame with lumped joint masses	506338.834586 rad²/s²	506331.91595 rad²/s²	+0.0014%
SR1 — Period — mode 1 · Response-spectrum analysis of a two-dimensional rigid frame	1.56213 s	1.562 s	+0.0080%
SR2 — Period — mode 1 · Response-spectrum analysis of a three-dimensional eccentric rigid-diaphragm frame	0.22705 s	0.2271 s	−0.021%
SR2b — Frequency — mode 1 · Response-spectrum analysis of a three-dimensional L-shaped braced frame	3.05908 Hz	3.0592 Hz	−0.0040%

Nonlinear Benchmarks

3 benchmarks

Benchmark	STRIX	Reference	Δ
SP1 — Tip Uz at Point2 (P=Mc/L, hardening peak) · Pushover cantilever moment hinge	24.9983 mm	24.998257 mm	+0.00017%
SH1 — Uniaxial backbone, cap boundary (kappa=thetaP) · DcrPMMHinge3d custom P-M-M column hinge — element-mechanics defense	1.25×10⁸ N·mm	1.25×10⁸ N·mm	0.00%
TH1 — Peak relative displacement, zeta=5%, dt=0.003125s (finest) · SDOF anchor for THA time integration	200.783 mm	200.786209 mm	−0.0016%

Stabilization Studies

1 benchmark

Benchmark	STRIX	Reference	Δ
P3S2 — Membrane · mode 1 (X 68%) · Custom stabilization parameter sensitivity	0.050509 s	0.050529 s	+0.0394%

Commercial cross-checks (MIDAS / ETABS)

Representative full-building results, cross-checked against commercial software as a secondary reference — supplementing, not replacing, the closed-form suite above.

Dynamic & modal

Quantity	STRIX	Reference	Δ
1st mode period (X)	3.01 s	3.14 s	−4.1%
1st mode period (Y)	2.58 s	2.69 s	−4.1%
Mass participation (X, 1st)	91.2%	90.4%	+0.9%

Member forces

Quantity	STRIX	Reference	Δ
Wall base moment	9,947 kN·m	9,606 kN·m	+3.5%
Wall axial Pu (1F)	502 kN	524 kN	−4.2%
Wall in-plane DL moment	1,180 kN·m	1,101 kN·m	+7.2%

Design results

Quantity	STRIX	Reference	Δ
Column capacity ratio (flexure)	1.77	1.74	+1.7%
Column capacity ratio (shear)	1.56	1.54	+1.3%
Wall flexural strength Mn	7,436 kN·m	7,388 kN·m	+0.6%

Representative full-building results from internal cross-verification models. Full per-case datasets and convergence plots for the closed-form suite are in the downloadable V&V report above.

Representative figures. References used for verification only — no competitor UI, table layout or content is reproduced.

Start building with STRIX today

Download for Windows Back to overview