Szczegóły publikacji
Opis bibliograficzny
Analysis of anomalies in the thermal-mechanical fatigue process using selected IT tools / K. Jaśkowiec, D. WILK-KOŁODZIEJCZYK, K. Nosarzewski // Archives of Foundry Engineering [Dokument elektroniczny]. — Czasopismo elektroniczne ; ISSN 2299-2944 . — Tytuł poprz.: Archiwum Odlewnictwa. — 2025 — vol. 25 iss. 4, s. 5–9. — Wymagania systemowe: Adobe Reader. — Bibliogr. s. 9, Abstr. — Publikacja dostępna online od: 2025-10-27
Autorzy (3)
- Jaśkowiec Krzysztof
- AGHWilk-Kołodziejczyk Dorota
- AGHNosarzewski Kacper
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 165408 |
|---|---|
| Data dodania do BaDAP | 2026-01-15 |
| Tekst źródłowy | URL |
| DOI | 10.24425/afe.2025.155373 |
| Rok publikacji | 2025 |
| Typ publikacji | artykuł w czasopiśmie |
| Otwarty dostęp |  |
| Creative Commons | |
| Czasopismo/seria | Archives of Foundry Engineering |
Abstract
The article uses the results obtained during the tests of a wide group of metal alloys using a device operating by the Coffin method. The measure of resistance to thermal-mechanical fatigue is the number of cycles that the sample withstands before a macrocrack occurs, at a fixed current and temperature range. The device offers the possibility of working in two modes of sample mounting. The first mode allows the sample to freely elongate parallel to its axis, while the second mounting mode limits this elongation by using a transducer. The aim of the publication is to present possible solutions for anomaly detection. Anomaly detection concerns traps that may occur during the measurement process. Advanced machine learning methods were used to analyze and detect anomalies in data regarding thermal fatigue resistance. Isolation Forest and One-Class SVM algorithms were used for anomaly detection, which allow for effective identification of unusual patterns in the data. The conducted research confirmed the usefulness of one of the selected methods in the process of anomaly identification using the example of elongation.
