Datasets:

aceengineer
/

digitalmodel-integrity

license: cc-by-4.0
pretty_name: digitalmodel — Pipeline & Riser Integrity
tags:
  - analysis-results
configs:
  - config_name: wall_thickness
    data_files: wall_thickness.parquet
  - config_name: wall_thickness_tension_moment
    data_files: wall_thickness_tension_moment.parquet

digitalmodel — Pipeline & Riser Integrity

Analysis/computation results published for query + visualization via the Hugging Face datasets-server API. Generated by scripts/hf/save_results_to_hf.py.

Tables

config	rows	cols
`wall_thickness`	369	15
`wall_thickness_tension_moment`	354609	16

Data-quality notes

Own computed analysis from the digitalmodel wall-thickness solver, reshaped to tidy rows by scripts/hf/build_capability_datasets.py. Reference pipe geometry only (323.9 mm OD, X65) - no client or project-specific data. wall_thickness: the original pressure-only sweep; per-check columns are sparse BY DESIGN because codes do not share a check set, so a null means the code defines no such check. wall_thickness_tension_moment (#1915): the same pipe swept over wall thickness x effective tension x bending moment across all nine codes, 354,609 rows; load ranges derived from capacity at the 18 mm mid-range wall (T_max = 0.6T_y = 4.6498 MN, M_max = 0.6M_y = 355.59 kN.m). It exists because the pressure-only sweep left the API-STD-2RD combined check at 0.0 everywhere; over these axes it spans 0 to 3.88 and GOVERNS at 30,855 points. Utilisation above 1.0 is retained deliberately. Screening only.

Provenance

Source repo: vamseeachanta/digitalmodel · algorithm: wall_thickness_screening+tension_moment_study
Schema version: 1.0.0
Source snapshots (sha256):
wall_thickness.json — sha256 8303ff5fdd91f15a…
wall_thickness_tension_moment.json — sha256 f86193595d099087…

Nulls are genuine missing values (non-finite floats were coerced to null). Load a table: pandas.read_parquet(hf_hub_download("aceengineer/digitalmodel-integrity", "<config>.parquet", repo_type="dataset")).