Datasets:
Append drill_samsung96k_1 (zeta=0) rao rows
Browse files- README.md +35 -9
- rao.parquet +2 -2
README.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
---
|
| 2 |
license: cc-by-4.0
|
| 3 |
-
pretty_name:
|
| 4 |
tags:
|
| 5 |
- analysis-results
|
| 6 |
configs:
|
|
@@ -10,37 +10,63 @@ configs:
|
|
| 10 |
data_files: rao.parquet
|
| 11 |
---
|
| 12 |
|
| 13 |
-
#
|
| 14 |
|
| 15 |
Analysis/computation results published for query + visualization via the Hugging Face
|
| 16 |
-
datasets-server API.
|
|
|
|
| 17 |
|
| 18 |
## Tables
|
| 19 |
|
| 20 |
| config | rows | cols |
|
| 21 |
|---|---|---|
|
| 22 |
| `hydrostatics` | 2 | 7 |
|
| 23 |
-
| `rao` |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
## Data-quality notes
|
| 26 |
|
| 27 |
-
**Screening-grade
|
|
|
|
|
|
|
| 28 |
|
| 29 |
- **Low-frequency artifact (flagged, not omitted):** in `rao`, heave & sway at freq ≤ 0.35 rad/s (T ≥ 18 s) in non-beam headings are physically implausible (heave ≈ 7–9 m/m where beam seas correctly show ≈ 1.1). 40 rows carry `reliability = "low_freq_artifact"`. Use freq ≥ 0.5 rad/s. Tracked: digitalmodel#1548.
|
| 30 |
- **Withheld table — `added_mass_damping`:** the source matrices declare units kg / kg·m² but magnitudes are consistent with te / te·m² (~1000× off, e.g. heave added-mass 58,447 vs ~1e8 kg expected). Withheld pending unit reconciliation. Tracked: digitalmodel#1550.
|
| 31 |
- **Withheld — operation-envelope (Hs/Tp) limits:** derived limits reach implausible values (~17 m Hs heavy-lift) and would mislead as operational limits. A class panel mesh is required first.
|
| 32 |
- **Roll at resonance is high** (23.5 deg/m at ζ=5 %, 19.2 at ζ=10 %) — a real, expected consequence of light damping, not an artifact.
|
| 33 |
|
| 34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
## Provenance
|
| 37 |
|
| 38 |
-
- Source repo: `digitalmodel`
|
| 39 |
- Schema version: `1.0.0`
|
| 40 |
- Source snapshots (sha256):
|
| 41 |
|
| 42 |
-
- `hydrostatics.parquet` — sha256 `3059b2cd485a3be5…`
|
| 43 |
-
- `rao.parquet` — sha256 `
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
|
| 45 |
Nulls are genuine missing values (non-finite floats were coerced to null). Load a table:
|
| 46 |
`pandas.read_parquet(hf_hub_download("aceengineer/digitalmodel-diffraction-rao", "<config>.parquet", repo_type="dataset"))`.
|
|
|
|
| 1 |
---
|
| 2 |
license: cc-by-4.0
|
| 3 |
+
pretty_name: digitalmodel Diffraction RAOs (multi-vessel, de-identified)
|
| 4 |
tags:
|
| 5 |
- analysis-results
|
| 6 |
configs:
|
|
|
|
| 10 |
data_files: rao.parquet
|
| 11 |
---
|
| 12 |
|
| 13 |
+
# digitalmodel Diffraction RAOs (multi-vessel, de-identified)
|
| 14 |
|
| 15 |
Analysis/computation results published for query + visualization via the Hugging Face
|
| 16 |
+
datasets-server API. One row per (vessel, damping, dof, frequency, heading). Vessels are
|
| 17 |
+
referenced only by de-identified vessel_db IDs.
|
| 18 |
|
| 19 |
## Tables
|
| 20 |
|
| 21 |
| config | rows | cols |
|
| 22 |
|---|---|---|
|
| 23 |
| `hydrostatics` | 2 | 7 |
|
| 24 |
+
| `rao` | 3300 | 10 |
|
| 25 |
+
|
| 26 |
+
## Vessels in `rao`
|
| 27 |
+
|
| 28 |
+
| vessel_id | solver | damping_zeta_pct | rows | basis |
|
| 29 |
+
|---|---|---|---|---|
|
| 30 |
+
| `install_bokalift_2` | OrcaWave | 5, 10 | 600 | parametric hull from public particulars |
|
| 31 |
+
| `drill_samsung96k_1` | AQWA | 0 | 2700 | hull-class-level de-identified model, survival loading condition |
|
| 32 |
|
| 33 |
## Data-quality notes
|
| 34 |
|
| 35 |
+
**Screening-grade — faithful-to-solver ≠ correct for design.** Neither model is class-approved.
|
| 36 |
+
|
| 37 |
+
### install_bokalift_2 (OrcaWave, parametric screening)
|
| 38 |
|
| 39 |
- **Low-frequency artifact (flagged, not omitted):** in `rao`, heave & sway at freq ≤ 0.35 rad/s (T ≥ 18 s) in non-beam headings are physically implausible (heave ≈ 7–9 m/m where beam seas correctly show ≈ 1.1). 40 rows carry `reliability = "low_freq_artifact"`. Use freq ≥ 0.5 rad/s. Tracked: digitalmodel#1548.
|
| 40 |
- **Withheld table — `added_mass_damping`:** the source matrices declare units kg / kg·m² but magnitudes are consistent with te / te·m² (~1000× off, e.g. heave added-mass 58,447 vs ~1e8 kg expected). Withheld pending unit reconciliation. Tracked: digitalmodel#1550.
|
| 41 |
- **Withheld — operation-envelope (Hs/Tp) limits:** derived limits reach implausible values (~17 m Hs heavy-lift) and would mislead as operational limits. A class panel mesh is required first.
|
| 42 |
- **Roll at resonance is high** (23.5 deg/m at ζ=5 %, 19.2 at ζ=10 %) — a real, expected consequence of light damping, not an artifact.
|
| 43 |
|
| 44 |
+
### drill_samsung96k_1 (AQWA, potential-flow)
|
| 45 |
+
|
| 46 |
+
- **`damping_zeta_pct = 0` means NO external/viscous roll damping** — radiation damping only.
|
| 47 |
+
The roll peak (15.6 deg/m at T = 19 s, beam seas) is therefore an undamped potential-flow
|
| 48 |
+
upper bound, not an operability prediction; apply an external roll damping model before use.
|
| 49 |
+
- **Frequency band is roll-focused:** 18 frequencies covering T ≈ 15.5–31.4 s (refined around
|
| 50 |
+
the measured roll-response band). No rows outside this band — absence is by scope, not omission.
|
| 51 |
+
- **Moonpool:** the panel model includes moonpool walls with no damping lid; moonpool-resonance
|
| 52 |
+
effects near the moonpool natural period are expected and not suppressed.
|
| 53 |
+
- Long-period sanity passed: beam-seas heave → 1.0 m/m and head-seas surge → 0.95 m/m at T = 31.4 s.
|
| 54 |
+
- **Withheld — added mass & damping matrices:** present in the source results but not yet
|
| 55 |
+
published pending the same unit reconciliation as digitalmodel#1550.
|
| 56 |
|
| 57 |
## Provenance
|
| 58 |
|
| 59 |
+
- Source repo: `digitalmodel` · solvers: `orcawave-diffraction@1537` (install_bokalift_2), AQWA v261 licensed-run lane (drill_samsung96k_1, run `lr_acma_bb24945c4b8b`)
|
| 60 |
- Schema version: `1.0.0`
|
| 61 |
- Source snapshots (sha256):
|
| 62 |
|
| 63 |
+
- `hydrostatics.parquet` — sha256 `3059b2cd485a3be5…` (unchanged)
|
| 64 |
+
- `rao.parquet` — sha256 `1a576f4c75059bdcaf90c3db58af33c1d408cb081e137e7bb3a945ba5b3f98f5`
|
| 65 |
+
- drill_samsung96k_1 de-identified source JSON — sha256 `ef6e7d90b52b4bf730c94e7aebcc1b295008ccb9b9c80ad3e0ef1a5d26c05dec`
|
| 66 |
+
|
| 67 |
+
Vessels are de-identified: `install_bokalift_2` from public particulars; `drill_samsung96k_1`
|
| 68 |
+
identified only at hull-class level. Not a run-identity ledger (that is the separate
|
| 69 |
+
`aceengineer/*-runs` contract).
|
| 70 |
|
| 71 |
Nulls are genuine missing values (non-finite floats were coerced to null). Load a table:
|
| 72 |
`pandas.read_parquet(hf_hub_download("aceengineer/digitalmodel-diffraction-rao", "<config>.parquet", repo_type="dataset"))`.
|
rao.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a576f4c75059bdcaf90c3db58af33c1d408cb081e137e7bb3a945ba5b3f98f5
|
| 3 |
+
size 48294
|