fix: critical audit fixes (C1-C4, H1-H5, M1)

C1: Fix 3 SyntaxError files (<REDACTED_SKIRT_L> -> float('nan'))
C2: Add NaN fail-fast guards in 6 files (raises RuntimeError)
C3: Remove phantom tutorial toctree (tutorials don't exist yet)
C4: Update all V&V counts: 39 benchmarks, 35/38 verified (92%)

H1: Fix config filename site_a_site.yaml -> site_a.yaml (16 refs)
H2: Clarify Zaaijer #10 note (different foundation types)
H4: Correct test counts: 140 tests across 15 modules (was 121/14)
H5: Add 10 missing modules to api.rst (fatigue, viz_*, encoder)

M1: np.trapz -> np.trapezoid (25 calls across 9 files)

Docs updated: README badge, VV_REPORT, cross_validation.rst,
verification.rst, environment.rst, CHANGELOG, api.rst, index.rst

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: Kyeong Sun Kim

CHANGELOG.md

@@ -92,7 +92,7 @@ development plan.
 
 ### Test totals
 
-- 121 active V&V tests (all PASS)
+- 140 active V&V tests (all PASS)
 - 33 example smoke tests across 11 turbines
 - 4 calibration regression tests against published references
 - 35/36 DNV-ST-0126 conformance checks

README.md

@@ -5,7 +5,7 @@
 [![Python 3.12](https://img.shields.io/badge/python-3.12-blue.svg)](https://www.python.org/downloads/release/python-3120/)
 [![OpenFAST](https://img.shields.io/badge/OpenFAST-v5.0.0-orange.svg)](https://github.com/OpenFAST/openfast/releases/tag/v5.0.0)
 [![OpenSeesPy](https://img.shields.io/badge/OpenSeesPy-3.7+-green.svg)](https://github.com/zhuminjie/OpenSeesPy)
-[![V&V](https://img.shields.io/badge/V%26V-27%2F28_benchmarks_(96%25)-brightgreen.svg)](validation/cross_validations/VV_REPORT.md)
+[![V&V](https://img.shields.io/badge/V%26V-35%2F38_benchmarks_(92%25)-brightgreen.svg)](validation/cross_validations/VV_REPORT.md)
 [![Tests](https://img.shields.io/badge/tests-45%20passing-brightgreen.svg)](op3_viz/tests/)
 [![Coverage](https://img.shields.io/badge/coverage-83.3%25-brightgreen.svg)](op3_viz/tests/)
 [![Version](https://img.shields.io/badge/version-1.0.0--rc1-blue.svg)](CHANGELOG.md)
@@ -87,18 +87,19 @@ python scripts/release_validation_report.py   # 18/19 PASS in ~42 s
 | Monte Carlo soil propagation | ✅ | ❌ | manual | manual | ❌ |
 | Hermite polynomial chaos expansion | ✅ | ❌ | ❌ | manual | ❌ |
 | Grid Bayesian calibration | ✅ | ❌ | ❌ | manual | ❌ |
-| V&V test suite | **121** | proprietary | proprietary | user-built | ~200 r-test |
+| V&V test suite | **140** | proprietary | proprietary | user-built | ~200 r-test |
 | License | Apache-2.0 | commercial | commercial | BSD-3 | Apache-2.0 |
 | Python-native | ✅ | ❌ | ❌ | wrapper | wrapper |
 
 ## v1.0.0-rc1 release highlights
 
-- **27 / 28 cross-validation benchmarks verified (96%)** against 20+
+- **35 / 38 cross-validation benchmarks verified (92%)** against 20+
   published sources (Fu & Bienen 2017, Vulpe 2015, Doherty 2005,
   Houlsby 2005, Jalbi 2018, Gazetas 2018, and 14 more)
-- **115+ unit tests pass** across 14 modules (code verification,
+- **140 unit tests pass** across 15 modules (code verification,
   consistency, sensitivity, extended invariants, PISA, cyclic
-  degradation, HSsmall, Mode D, UQ, reproducibility snapshot, ...)
+  degradation, HSsmall, Mode D, UQ, reproducibility snapshot,
+  framework integration, ...)
 - **4 / 4 calibration regression** against published references
   with all four examples within 4% of the most stringent
   (NREL 5 MW OC3 at **-0.4%** vs Jonkman & Musial 2010)
@@ -121,8 +122,8 @@ python scripts/release_validation_report.py   # 18/19 PASS in ~42 s
 
 ## Cross-Validation Against Published Benchmarks
 
-Op3 has been cross-validated against **31 independent benchmarks** from
-20+ published sources. **27 of 28 in-scope benchmarks verified (96%).**
+Op3 has been cross-validated against **39 independent benchmarks** from
+20+ published sources. **35 of 38 in-scope benchmarks verified (92%).**
 
 | Category | Benchmarks | Error range | Sources |
 |---|---|---|---|
@@ -135,6 +136,13 @@ Op3 has been cross-validated against **31 independent benchmarks** from
 | PISA clay stiffness | #6 | 16--32% | Burd et al. 2020 |
 | VH envelope | #8 | -7.7% | Houlsby & Byrne / Vulpe 2015 |
 | p_ult(z) profile | #21 | consistent | OptumGX plate extraction |
+| Centrifuge yield moment | #22 | **-0.7%** | DJ Kim et al. 2014 |
+| Full-scale tripod f1 | #24 | -0.2% | Seo et al. 2020 |
+| Walney 1 monopile f1 | #25 | -2.1% | Arany et al. 2015 |
+| Suction bucket scour sensitivity | #26 | within range | Cheng et al. 2024 |
+| f_meas/f_design ratio | #27 | +0.3% | Kallehave et al. 2015 |
+| Cyclic rotation (N=100, N=1M) | #28 | 3.7--4.3% | Jeong et al. 2021 |
+| OC4 jacket f1 (fixed-base) | #29 | +1.9% | Popko et al. 2012 |
 
 Key results:
 - **NcV = 6.006** (ref 5.94, +1.1%) -- textbook bearing capacity match
@@ -218,7 +226,7 @@ op3/                     the Python package
   uq/                    Propagation / PCE / Bayesian
   sacs_interface/        SACS jacket deck parser
 
-tests/                   121 active V&V tests
+tests/                   140 active V&V tests
 scripts/                 Runners, audits, regressions, release tooling
 examples/                11 turbine TowerModel build.py files
 docs/                    Sphinx + tutorials + Mode D notes + developer notes
@@ -355,7 +363,7 @@ modes — only the foundation boundary condition changes.
 |  D   | **Dissipation-weighted generalized BNWF** | High | Slow | Full energy-consistent coupling with OptumGX plastic dissipation field |
 
 All four modes share the same tower, rotor, and nacelle inertia
-properties from [`op3/config/site_a_site.yaml`](op3/config/site_a_site.yaml).
+properties from [`op3/config/site_a.yaml`](op3/config/site_a.yaml).
 Only the foundation representation changes, which makes it trivial to
 compare the effect of foundation modeling choice on the predicted
 natural frequency, mode shape, or transient response.
@@ -476,7 +484,7 @@ The extraction is handled by
 The SubDyn input file is generated by
 [`op3/openfast_coupling/build_site_a_subdyn.py`](op3/openfast_coupling/build_site_a_subdyn.py).
 Both scripts are driven by the single-source-of-truth YAML at
-[`op3/config/site_a_site.yaml`](op3/config/site_a_site.yaml).
+[`op3/config/site_a.yaml`](op3/config/site_a.yaml).
 
 ## NREL reference library bundled in this repository
 
@@ -560,7 +568,7 @@ numerical_model/
 │   ├── openfast_coupling/             OpenSees ↔ SubDyn bridge
 │   ├── integration/                   OptumGX outputs → OpenSees springs
 │   └── config/                        single source of truth
-│       └── site_a_site.yaml
+│       └── site_a.yaml
 │
 ├── data/                              OptumGX persisted outputs
 │   ├── integrated_database_1794.csv   master Monte Carlo database

docs/FRAMEWORK.md

@@ -311,7 +311,7 @@ profile, material properties, environmental conditions — lives in
 exactly one file:
 
 ```
-op3/config/site_a_site.yaml
+op3/config/site_a.yaml
 ```
 
 This is a strict convention. No constants are hardcoded in any
@@ -320,7 +320,7 @@ SSOT YAML via:
 
 ```python
 from op3.config import load_site_config
-cfg = load_site_config('op3/config/site_a_site.yaml')
+cfg = load_site_config('op3/config/site_a.yaml')
 D = cfg['foundation']['bucket_diameter_m']
 L = cfg['foundation']['skirt_length_m']
 ```
@@ -369,7 +369,7 @@ The framework is designed so that porting to a new turbine touches
 only the SSOT configuration and the OptumGX upstream step. Specific
 steps:
 
-1. Create a new YAML file modeled on `op3/config/site_a_site.yaml`
+1. Create a new YAML file modeled on `op3/config/site_a.yaml`
    with the new turbine's geometry, soil profile, rotor mass, tower
    properties, and water depth.
 2. Run OptumGX over the new parameter envelope (requires a license)

docs/OPTUMGX_BOUNDARY.md

@@ -199,7 +199,7 @@ If you have an OptumGX academic license and the `optumgx` Python API:
    profile. A license holder can swap in a different constitutive
    model (Mohr-Coulomb, Hoek-Brown, user-defined) and regenerate
    the capacity envelope.
-3. **Apply to a different site.** Replace `config/site_a_site.yaml`
+3. **Apply to a different site.** Replace `config/site_a.yaml`
    with a different site's geometry and soil, rerun OptumGX, and
    the downstream OpenSeesPy and OpenFAST pipelines pick up the new
    CSVs automatically.

docs/sphinx/api.rst

@@ -67,6 +67,33 @@ OpenFAST coupling
 .. automodule:: op3.openfast_coupling.soildyn_export
    :members:
 
+Fatigue
+-------
+
+.. automodule:: op3.fatigue
+   :members:
+
+Visualization
+-------------
+
+.. automodule:: op3.visualization
+   :members:
+
+.. automodule:: op3.viz_optumgx
+   :members:
+
+.. automodule:: op3.viz_openfast
+   :members:
+
+.. automodule:: op3.viz_tier1
+   :members:
+
+.. automodule:: op3.viz_tier2
+   :members:
+
+.. automodule:: op3.viz_tier3
+   :members:
+
 Uncertainty quantification (Phase 5)
 ------------------------------------
 
@@ -78,3 +105,15 @@ Uncertainty quantification (Phase 5)
 
 .. automodule:: op3.uq.bayesian
    :members:
+
+Advanced UQ
+-----------
+
+.. automodule:: op3.uq.sequential_bayesian
+   :members:
+
+.. automodule:: op3.uq.encoder_bridge
+   :members:
+
+.. automodule:: op3.uq.encoder_training
+   :members:

docs/sphinx/cross_validation.rst

@@ -1,12 +1,12 @@
 Cross-Validation Against Published Benchmarks
 ==============================================
 
-Op\ :sup:`3` v1.0 has been cross-validated against 31 independent
+Op\ :sup:`3` v1.0 has been cross-validated against 39 independent
 benchmarks drawn from 20+ published sources spanning centrifuge
 experiments, field trials, 3D finite-element analyses, closed-form
 analytical solutions, and design code requirements.
 
-**Overall score: 27 of 28 in-scope benchmarks verified (96%).**
+**Overall score: 35 of 38 in-scope benchmarks verified (92%).**
 
 .. contents:: On this page
    :local:
@@ -135,6 +135,48 @@ Summary table
      - depth profile
      - consistent
      - verified
+   * - 22
+     - DJ Kim tripod M\ :sub:`y` at yield
+     - DJ Kim et al. (2014)
+     - M\ :sub:`y` (MNm)
+     - -0.7%
+     - verified
+   * - 24
+     - Seo 2020 full-scale tripod f\ :sub:`1`
+     - Seo et al. (2020)
+     - f\ :sub:`1` (Hz)
+     - -0.2%
+     - verified
+   * - 25
+     - Arany Walney 1 f\ :sub:`1`
+     - Arany et al. (2015)
+     - f\ :sub:`1` (Hz)
+     - -2.1%
+     - verified
+   * - 26
+     - Cheng 2024 scour df/f\ :sub:`0`
+     - Cheng et al. (2024)
+     - df/f\ :sub:`0` (%)
+     - -40% (both <1%)
+     - verified
+   * - 27
+     - Kallehave f\ :sub:`meas`/f\ :sub:`design`
+     - Kallehave et al. (2015)
+     - ratio
+     - +0.3%
+     - verified
+   * - 28
+     - Jeong 2021 cyclic rotation
+     - Jeong et al. (2021)
+     - rotation (deg)
+     - 3.7--4.3%
+     - verified
+   * - 29
+     - OC4 jacket f\ :sub:`1` (fixed-base)
+     - Popko et al. (2012)
+     - f\ :sub:`1` (Hz)
+     - +1.9%
+     - verified
    * - 7
      - PISA Dunkirk sand
      - Byrne et al. (2020)
@@ -402,8 +444,8 @@ Reference data
 
 All reference data is stored in machine-readable format:
 
-* ``validation/cross_validations/extended_reference_data.py`` -- 19
-  Python dictionaries covering 19 published sources
+* ``validation/cross_validations/extended_reference_data.py`` -- 20+
+  Python dictionaries covering 20+ published sources
 * ``validation/cross_validations/extracted_benchmark_data.json`` -- 36
   individual benchmark entries
 * ``validation/cross_validations/all_results.json`` -- consolidated

docs/sphinx/environment.rst

@@ -3,7 +3,7 @@ Environment setup guide
 
 This page is the complete setup guide for running Op\ :sup:`3` on a
 fresh machine. Follow it top-to-bottom and you will have a working
-installation with all 121 V&V tests passing in under 30 minutes.
+installation with all 140 V&V tests passing in under 30 minutes.
 
 Supported platforms
 -------------------

docs/sphinx/index.rst

@@ -42,16 +42,12 @@ Foundations"* (Seoul National University, June 2026).
    verification
    cross_validation
 
-.. toctree::
-   :maxdepth: 2
-   :caption: Tutorials
+.. note::
 
-   tutorials/01_quickstart
-   tutorials/02_foundation_modes
-   tutorials/03_uncertainty_quantification
-   tutorials/04_calibration
-   tutorials/05_soildyn_export
-   tutorials/06_dlc_sweep
+   **Tutorials** are planned for the v1.1 release. In the meantime,
+   see the demo scripts in ``scripts/`` (e.g., ``visualize_op3.py``,
+   ``visualize_optumgx_openfast.py``, ``test_three_analyses.py``)
+   and the sample projects in ``sample_projects/``.
 
 .. toctree::
    :maxdepth: 2
@@ -90,11 +86,11 @@ Run the bundled examples
 Verification & validation
 -------------------------
 
-The full V&V suite consists of 115 active tests across 14 modules
-plus 31 cross-validation benchmarks against 20+ published sources.
+The full V&V suite consists of 140 active tests across 15 modules
+plus 39 cross-validation benchmarks against 20+ published sources.
 See :doc:`verification` for the unit-test falsification gates and
 :doc:`cross_validation` for the published-benchmark comparison
-(27/28 in-scope benchmarks verified, 96%).
+(35/38 in-scope benchmarks verified, 92%).
 
 Indices and tables
 ==================

docs/sphinx/verification.rst

@@ -1,8 +1,8 @@
 Verification & Validation
 =========================
 
-The Op\ :sup:`3` V&V suite contains 115+ active falsification gates
-across 14 test modules. All tests must pass before any commit reaches
+The Op\ :sup:`3` V&V suite contains 140 active falsification gates
+across 15 test modules. All tests must pass before any commit reaches
 ``main``.
 
 Test summary
@@ -24,7 +24,7 @@ Test summary
      - 5
      - Physical sensitivity invariants
    * - ``tests/test_extended_vv.py``
-     - 8
+     - 12
      - Damping, energy, reciprocity, units, coordinates, orthogonality
    * - ``tests/test_pisa.py``
      - 9
@@ -47,6 +47,9 @@ Test summary
    * - ``tests/test_uq.py``
      - 13
      - MC propagation, PCE, Bayesian
+   * - ``tests/test_op3_framework.py``
+     - 15
+     - Framework integration, project I/O, visualization wiring
    * - ``tests/test_reproducibility.py``
      - 6
      - Snapshot byte-identical reproduction
@@ -104,8 +107,8 @@ Cross-validation against published benchmarks
 ----------------------------------------------
 
 In addition to the internal V&V tests above, Op\ :sup:`3` has been
-cross-validated against **31 independent benchmarks** from 20+ published
-sources, achieving a **96% verification rate** (27/28 in-scope).
+cross-validated against **39 independent benchmarks** from 20+ published
+sources, achieving a **92% verification rate** (35/38 in-scope).
 
 See :doc:`cross_validation` for the complete cross-validation report
 including OptumGX bearing capacity factors (0.8--7.8% error vs Vulpe

op3/config.py

@@ -8,7 +8,7 @@
 Usage:
 
     from op3 import load_site_config
-    cfg = load_site_config('op3/config/site_a_site.yaml')
+    cfg = load_site_config('op3/config/site_a.yaml')
     D = cfg['foundation']['bucket_diameter_m']
 """
 from __future__ import annotations