Vibration data from a scaled railway axle mounted on a laboratory test rig designed to reproduce controlled rotating operating conditions. The dataset, organized by CASTEJON, CRISTINA, compares healthy operation to a scenario with intentionally introduced mechanical looseness. Data is grouped into folders for healthy ('EJE SANO') and faulty ('HOLGURA') conditions, with subfolders for rotation speeds of 20, 40, and 60 Hz.
Use Cases
- Train fault detection models based on vibration data from a rotor with mechanical looseness.
- Compare vibration signatures between healthy and faulty machinery operating at different speeds.
- Benchmark signal processing techniques for structural health monitoring using controlled laboratory data.
- Study the effects of mechanical looseness in shaft-bearing and motor-coupling interfaces on vibration patterns.
Strengths
- Data is organized into distinct folders for healthy and faulty conditions, enabling direct comparison.
- Experiments were conducted at controlled rotation speeds (20, 40, 60 Hz), providing a structured test environment.
- Mechanical looseness was intentionally introduced in specific assembly interfaces, creating a clear fault scenario.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count and file size are unknown, which may limit suitability assessment for large-scale training.
Provenance
- Source
- e-cienciaDatos Harvested Dataverse
- Collection Method
- Collected from a laboratory test rig (Rotokit of SpectraQuest) with a scaled railway axle under controlled conditions.
- Freshness
- Last updated 2026-05-31 04:10:15; freshness should be verified.