Preprocessing of Inconsistent Creep Data Collected from a Literature Survey to Provide Reliable and Consistent Creep Life Prediction
Source abstract
Abstract
Numerous studies on machine learning-based creep life prediction using the creep data collected from various existing experimental data have been reported. Since the prediction of creep life is heavily influenced by the quality and integrity of the collected creep data, data preprocessing is required to eliminate physically inconsistent creep datapoints, known as outliers. In the present study, a machine learning-based data screening methodology was developed to detect and eliminate outliers from the creep data collected from a survey of various studies in the literature. The methodology consisted of selecting appropriate machine learning models for the collected creep data through an assessment of their validity in creep physics, evaluating the prediction accuracy and variability of collected datapoints through bagging of the selected machine learning models, and identifying inconsistent datapoints by ranking their residuals and prediction variabilities. The proposed methodology for detecting and eliminating outliers was successfully applied to the multi-source collected creep data of a Ni-base single crystal superalloy CMSX-4 and led to improved accuracy and consistency in creep life prediction. In addition, the proposed methodology was validated by predicting the creep life of a newly generated creep dataset that was not exposed to any model training using a machine learning model trained and optimized by the outlier-eliminated creep data.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Fatigue and fracture mechanics · High Temperature Alloys and Creep · Probabilistic and Robust Engineering Design
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (39)
Show source-provided references
1. 10.3365/kjmm.2023.61.4.30110.3365/kjmm.2023.61.4.301 ↗
2. 10.3365/kjmm.2024.62.10.76610.3365/kjmm.2024.62.10.766 ↗
3. Monkman · 1956 · 593
4. 10.1016/s0079-6425(99)00006-710.1016/s0079-6425(99)00006-7 ↗
5. 10.1016/j.msea.2006.08.09410.1016/j.msea.2006.08.094 ↗
6. Larson · 1952 · 765
7. 10.1016/j.scriptamat.2006.12.03310.1016/j.scriptamat.2006.12.033 ↗
8. 10.1179/026708310x1271241031173010.1179/026708310x12712410311730 ↗
9. 10.1016/j.matdes.2021.11032610.1016/j.matdes.2021.110326 ↗
10. 10.1016/j.jmrt.2021.04.07910.1016/j.jmrt.2021.04.079 ↗
11. 10.1016/j.actamat.2020.05.00110.1016/j.actamat.2020.05.001 ↗
12. 10.1016/j.actamat.2019.02.01710.1016/j.actamat.2019.02.017 ↗
13. 10.1007/s12540-023-01445-310.1007/s12540-023-01445-3 ↗
14. 10.1007/s12540-023-01531-610.1007/s12540-023-01531-6 ↗
15. Kong · 2022 · 166123
16. Murakami · 1999 · 1143
17. Jiang · 2003 · 3871
18. Evans · 1988 · 2187
19. Kitaguchi · 2020 · 15314
20. Miller · 2000 · 1
21. Maruyama · 2021 · 113699
22. Wang · 2021 · 1
23. Laughlin · 2014 · 1
24. Morinaga · 2010 · 1
25. Davis · 2000 · 1
26. Nitta · 2023 · 107797
27. Padilha · 2003 · 159
28. Kim · 2019 · 145
29. Allen · 1990 · 1937
30. 10.1016/j.ultramic.2011.01.02710.1016/j.ultramic.2011.01.027 ↗
31. 10.1016/j.corcom.2021.06.00310.1016/j.corcom.2021.06.003 ↗
32. Rösler · 2005 · 1397
33. Kassner · 2015 · 1
34. 10.1016/s1359-6454(99)00217-710.1016/s1359-6454(99)00217-7 ↗
35. Rösler · 2010 · 1815
36. 10.1016/j.actamat.2012.01.03910.1016/j.actamat.2012.01.039 ↗
37. 10.1016/j.actamat.2006.09.02610.1016/j.actamat.2006.09.026 ↗
38. 10.1016/j.actamat.2005.11.01610.1016/j.actamat.2005.11.016 ↗
39. Zhu · 2013 · 374