Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty
Additive manufacturing (AM) is known for versatile fabrication of complex parts, while also allowing the synthesis of materials with desired microstructures and resulting properties. These benefits come at a cost: process control to manufacture parts within given specifications is very challenging d...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-03-01
|
| Series: | Materials & Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0264127523001144 |
| _version_ | 1797854026173251584 |
|---|---|
| author | Dmitry Chernyavsky Denys Y. Kononenko Jun Hee Han Hwi Jun Kim Jeroen van den Brink Konrad Kosiba |
| author_facet | Dmitry Chernyavsky Denys Y. Kononenko Jun Hee Han Hwi Jun Kim Jeroen van den Brink Konrad Kosiba |
| author_sort | Dmitry Chernyavsky |
| collection | DOAJ |
| description | Additive manufacturing (AM) is known for versatile fabrication of complex parts, while also allowing the synthesis of materials with desired microstructures and resulting properties. These benefits come at a cost: process control to manufacture parts within given specifications is very challenging due to the relevance of a large number of processing parameters. Efficient predictive machine learning (ML) models trained on small datasets, can minimize this cost. They also allow to assess the quality of the dataset inclusive of uncertainty. This is important in order for additively manufactured parts to meet property specifications not only on average, but also within a given variance or uncertainty. Here, we demonstrate this strategy by developing a heteroscedastic Gaussian process (HGP) model, from a dataset based on laser powder bed fusion of a glass-forming alloy at varying processing parameters. Using amorphicity as the microstructural descriptor, we train the model on our Zr52.5Cu17.9Ni14.6Al10Ti5 (at.%) alloy dataset. The HGP model not only accurately predicts the mean value of amorphicity, but also provides the respective uncertainty. The quantification of the aleatoric and epistemic uncertainty contributions allows to assess intrinsic inaccuracies of the dataset, as well as identify underlying physical phenomena. This HGP model approach enables to systematically improve ML-driven AM processes. |
| first_indexed | 2024-04-09T19:59:11Z |
| format | Article |
| id | doaj.art-f5c092a2c23d4238a57e1fc8ba0b7cb9 |
| institution | Directory Open Access Journal |
| issn | 0264-1275 |
| language | English |
| last_indexed | 2024-04-09T19:59:11Z |
| publishDate | 2023-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Materials & Design |
| spelling | doaj.art-f5c092a2c23d4238a57e1fc8ba0b7cb92023-04-03T05:20:41ZengElsevierMaterials & Design0264-12752023-03-01227111699Machine learning for additive manufacturing: Predicting materials characteristics and their uncertaintyDmitry Chernyavsky0Denys Y. Kononenko1Jun Hee Han2Hwi Jun Kim3Jeroen van den Brink4Konrad Kosiba5Leibniz Institute of Solid State and Materials Science (IFW Dresden), Helmholtzstr.20, 01069 Dresden, GermanyLeibniz Institute of Solid State and Materials Science (IFW Dresden), Helmholtzstr.20, 01069 Dresden, GermanyKorea Institute of Industrial Technology (KITECH), Korea Institute for Rare Metals, 156, Gaetbeol-ro, Yeonsu-gu, 21999 Incheon, South KoreaKorea Institute of Industrial Technology (KITECH), Korea Institute for Rare Metals, 156, Gaetbeol-ro, Yeonsu-gu, 21999 Incheon, South KoreaLeibniz Institute of Solid State and Materials Science (IFW Dresden), Helmholtzstr.20, 01069 Dresden, GermanyLeibniz Institute of Solid State and Materials Science (IFW Dresden), Helmholtzstr.20, 01069 Dresden, GermanyAdditive manufacturing (AM) is known for versatile fabrication of complex parts, while also allowing the synthesis of materials with desired microstructures and resulting properties. These benefits come at a cost: process control to manufacture parts within given specifications is very challenging due to the relevance of a large number of processing parameters. Efficient predictive machine learning (ML) models trained on small datasets, can minimize this cost. They also allow to assess the quality of the dataset inclusive of uncertainty. This is important in order for additively manufactured parts to meet property specifications not only on average, but also within a given variance or uncertainty. Here, we demonstrate this strategy by developing a heteroscedastic Gaussian process (HGP) model, from a dataset based on laser powder bed fusion of a glass-forming alloy at varying processing parameters. Using amorphicity as the microstructural descriptor, we train the model on our Zr52.5Cu17.9Ni14.6Al10Ti5 (at.%) alloy dataset. The HGP model not only accurately predicts the mean value of amorphicity, but also provides the respective uncertainty. The quantification of the aleatoric and epistemic uncertainty contributions allows to assess intrinsic inaccuracies of the dataset, as well as identify underlying physical phenomena. This HGP model approach enables to systematically improve ML-driven AM processes.http://www.sciencedirect.com/science/article/pii/S0264127523001144Additive manufacturingLaser powder bed fusionMachine learningGaussian processesMetallic glassUncertainty quantification |
| spellingShingle | Dmitry Chernyavsky Denys Y. Kononenko Jun Hee Han Hwi Jun Kim Jeroen van den Brink Konrad Kosiba Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty Materials & Design Additive manufacturing Laser powder bed fusion Machine learning Gaussian processes Metallic glass Uncertainty quantification |
| title | Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty |
| title_full | Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty |
| title_fullStr | Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty |
| title_full_unstemmed | Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty |
| title_short | Machine learning for additive manufacturing: Predicting materials characteristics and their uncertainty |
| title_sort | machine learning for additive manufacturing predicting materials characteristics and their uncertainty |
| topic | Additive manufacturing Laser powder bed fusion Machine learning Gaussian processes Metallic glass Uncertainty quantification |
| url | http://www.sciencedirect.com/science/article/pii/S0264127523001144 |
| work_keys_str_mv | AT dmitrychernyavsky machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty AT denysykononenko machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty AT junheehan machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty AT hwijunkim machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty AT jeroenvandenbrink machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty AT konradkosiba machinelearningforadditivemanufacturingpredictingmaterialscharacteristicsandtheiruncertainty |