A generalized machine learning framework to estimate fatigue life across materials with minimal data
In this research, a generalized machine learning (ML) framework is proposed to estimate the fatigue life of epoxy polymers and additively manufactured AlSi10Mg alloy materials, leveraging their failure surface void characteristics. An extreme gradient boosting algorithm-based ML framework encompassi...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-10-01
|
| Series: | Materials & Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0264127524007305 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850266130381275136 |
|---|---|
| author | Dharun Vadugappatty Srinivasan Morteza Moradi Panagiotis Komninos Dimitrios Zarouchas Anastasios P. Vassilopoulos |
| author_facet | Dharun Vadugappatty Srinivasan Morteza Moradi Panagiotis Komninos Dimitrios Zarouchas Anastasios P. Vassilopoulos |
| author_sort | Dharun Vadugappatty Srinivasan |
| collection | DOAJ |
| description | In this research, a generalized machine learning (ML) framework is proposed to estimate the fatigue life of epoxy polymers and additively manufactured AlSi10Mg alloy materials, leveraging their failure surface void characteristics. An extreme gradient boosting algorithm-based ML framework encompassing Synthetic Minority Over-sampling TEchnique (SMOTE), categorical data encoding, and external loop cross-validation is developed to evaluate the fatigue life across materials. The influence of different training strategies based on materials, input features, encoding method, and data standardization on the model performance is explored. Additionally, the importance of anti-data-leakage and anti-overfitting measures over the ML model performance is addressed. The result shows that the data-leakage-free, external loop cross-validated model can estimate the fatigue life of selective epoxy polymers and metal alloys with an average R2 of 0.71 ± 0.06 using a mere 12 to 27 experimental data points per material category. Whereas the model trained with data-leakage and overfitting results in high R2 of 0.9. |
| format | Article |
| id | doaj-art-a038cd72b0fe4e93a55c86127e6ef19b |
| institution | OA Journals |
| issn | 0264-1275 |
| language | English |
| publishDate | 2024-10-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Materials & Design |
| spelling | doaj-art-a038cd72b0fe4e93a55c86127e6ef19b2025-08-20T01:54:15ZengElsevierMaterials & Design0264-12752024-10-0124611335510.1016/j.matdes.2024.113355A generalized machine learning framework to estimate fatigue life across materials with minimal dataDharun Vadugappatty Srinivasan0Morteza Moradi1Panagiotis Komninos2Dimitrios Zarouchas3Anastasios P. Vassilopoulos4Composite Mechanics Group (GR-MeC), Ecole Polytechnique Fédérale de Lausanne (EPFL), Station 16, CH-1015 Lausanne, Switzerland; Corresponding author.Center of Excellence in Artificial Intelligence for Structures, Prognostics & Health Management, Aerospace Engineering Faculty, Delft University of Technology, Kluyverweg 1, Delft 2629 HS, the NetherlandsCenter of Excellence in Artificial Intelligence for Structures, Prognostics & Health Management, Aerospace Engineering Faculty, Delft University of Technology, Kluyverweg 1, Delft 2629 HS, the NetherlandsCenter of Excellence in Artificial Intelligence for Structures, Prognostics & Health Management, Aerospace Engineering Faculty, Delft University of Technology, Kluyverweg 1, Delft 2629 HS, the NetherlandsComposite Mechanics Group (GR-MeC), Ecole Polytechnique Fédérale de Lausanne (EPFL), Station 16, CH-1015 Lausanne, SwitzerlandIn this research, a generalized machine learning (ML) framework is proposed to estimate the fatigue life of epoxy polymers and additively manufactured AlSi10Mg alloy materials, leveraging their failure surface void characteristics. An extreme gradient boosting algorithm-based ML framework encompassing Synthetic Minority Over-sampling TEchnique (SMOTE), categorical data encoding, and external loop cross-validation is developed to evaluate the fatigue life across materials. The influence of different training strategies based on materials, input features, encoding method, and data standardization on the model performance is explored. Additionally, the importance of anti-data-leakage and anti-overfitting measures over the ML model performance is addressed. The result shows that the data-leakage-free, external loop cross-validated model can estimate the fatigue life of selective epoxy polymers and metal alloys with an average R2 of 0.71 ± 0.06 using a mere 12 to 27 experimental data points per material category. Whereas the model trained with data-leakage and overfitting results in high R2 of 0.9.http://www.sciencedirect.com/science/article/pii/S0264127524007305CompositesFatigueMetal alloysMachine learningVoidMinimal data |
| spellingShingle | Dharun Vadugappatty Srinivasan Morteza Moradi Panagiotis Komninos Dimitrios Zarouchas Anastasios P. Vassilopoulos A generalized machine learning framework to estimate fatigue life across materials with minimal data Materials & Design Composites Fatigue Metal alloys Machine learning Void Minimal data |
| title | A generalized machine learning framework to estimate fatigue life across materials with minimal data |
| title_full | A generalized machine learning framework to estimate fatigue life across materials with minimal data |
| title_fullStr | A generalized machine learning framework to estimate fatigue life across materials with minimal data |
| title_full_unstemmed | A generalized machine learning framework to estimate fatigue life across materials with minimal data |
| title_short | A generalized machine learning framework to estimate fatigue life across materials with minimal data |
| title_sort | generalized machine learning framework to estimate fatigue life across materials with minimal data |
| topic | Composites Fatigue Metal alloys Machine learning Void Minimal data |
| url | http://www.sciencedirect.com/science/article/pii/S0264127524007305 |
| work_keys_str_mv | AT dharunvadugappattysrinivasan ageneralizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT mortezamoradi ageneralizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT panagiotiskomninos ageneralizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT dimitrioszarouchas ageneralizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT anastasiospvassilopoulos ageneralizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT dharunvadugappattysrinivasan generalizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT mortezamoradi generalizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT panagiotiskomninos generalizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT dimitrioszarouchas generalizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata AT anastasiospvassilopoulos generalizedmachinelearningframeworktoestimatefatiguelifeacrossmaterialswithminimaldata |